novox/mesh-host
8
0
Fork 0
Code Issues Pull Requests 1 Releases Wiki Activity

Compare commits

merge into: novox/mesh-host:57-config-and-provisioning
Branches Tags
novox/mesh-host:main novox/mesh-host:57-config-and-provisioning novox/mesh-host:feat/57-config-and-provisioning novox/mesh-host:replay/r92-before novox/mesh-host:feat/325-identities-and-kept-folders novox/mesh-host:fix/567-an-access-is-judged-by-who-writes-it novox/mesh-host:feat/the-operators-machine novox/mesh-host:fix/inspect-by-kind-not-the-ambiguous-bare-form
..
pull from: novox/mesh-host:replay/r92-before
Branches Tags
novox/mesh-host:57-config-and-provisioning novox/mesh-host:main novox/mesh-host:feat/57-config-and-provisioning novox/mesh-host:replay/r92-before novox/mesh-host:feat/325-identities-and-kept-folders novox/mesh-host:fix/567-an-access-is-judged-by-who-writes-it novox/mesh-host:feat/the-operators-machine novox/mesh-host:fix/inspect-by-kind-not-the-ambiguous-bare-form
1 Commits
57-config-and-provisioning .. replay/r92-before
Author SHA1 Message Date
jochen 9e233e1d87 Replay R92's commit before its fix: mesh-host main without genesis's registry trust (#91 reverted) 
Never to be merged. The replay of hq issue 92 runs genesis at the commit before its fix, and the fix's
own first parent stops at step 6 on issue 146; this is today's genesis without #91, so the bed
reaches the push it replays (mesh-lab replays/register.go R92).
 2026-10-13 02:57:43 +02:00
77 changed files with 231 additions and 7008 deletions
Expand all files Collapse all files
RELEASE-57.md
-6
View File
@@ -1,6 +0,0 @@
# Release 57: config and provisioning
This branch carries the work of hq design 57 (ADR 0348, ADR 0349) in this repository. Each package
merges here through its own reviewed pull request; the branch reaches main as one pull request, in one
delivery group with the branch of the same name in mesh-controller, mesh-host, mesh-sdk, mesh-catalog
and hq, after a rehearsal in the lab. This file is deleted by the last commit before that pull request.
cmd/mesh-bootstrap/main.go
-6
View File
@@ -116,8 +116,6 @@ const usage = `mesh-bootstrap — make a bare machine into a mesh
--tools-source the repository the shared base is built from
--tools-ref what of it to build (default main)
--tools-go-source the repository the Go toolchain is built from (default: mesh-tools-go beside --tools-source)
--tools-go-ref what of it to build (default main)
--catalog-source the catalogue REPOSITORY, for building its modules;
--catalog is the checkout that says what they are
--catalog-ref what of it to build (default main)
@@ -310,10 +308,6 @@ func newFlagSet(opts *bootstrap.Options, jsonOut *bool) *flag.FlagSet {
"the repository the shared base is built from")
set.StringVar(&opts.ToolsSource.Ref, "tools-ref", opts.ToolsSource.Ref,
"what of it to build (default main)")
set.StringVar(&opts.ToolsGoSource.Repository, "tools-go-source", opts.ToolsGoSource.Repository,
"the repository the Go toolchain is built from (default: mesh-tools-go beside --tools-source)")
set.StringVar(&opts.ToolsGoSource.Ref, "tools-go-ref", opts.ToolsGoSource.Ref,
"what of it to build (default main)")
set.StringVar(&opts.CatalogSource.Repository, "catalog-source", opts.CatalogSource.Repository,
"the catalogue REPOSITORY, for building its modules — --catalog is the checkout that says what they are")
set.StringVar(&opts.CatalogSource.Ref, "catalog-ref", opts.CatalogSource.Ref,
cmd/mesh-host/busy_test.go
-66
View File
@@ -1,66 +0,0 @@
package main
import (
"context"
"sync/atomic"
"testing"
"time"
"github.com/novox/mesh-host/internal/link"
)
// A container waiting because it said busy is news when it starts waiting and when it is replaced, never on
// its output alone (novox/hq ADR 0335): the walk waiting on this machine starts its gate at the report that
// says it replaced, and a reconcile is otherwise silent.
func TestABusyWaitIsSaidWhenItStartsAndWhenItEnds(t *testing.T) {
quiet := link.Report{Outward: []string{"eth0"}}
waiting := link.Report{Outward: []string{"eth0"}, Busy: []link.Busy{{ID: "build-agent.server",
Container: "mesh-build-agent", State: "waiting", Said: "building a"}}}
still := link.Report{Outward: []string{"eth0"}, Busy: []link.Busy{{ID: "build-agent.server",
Container: "mesh-build-agent", State: "waiting", Said: "building b, a minute on"}}}
replaced := link.Report{Outward: []string{"eth0"}, Busy: []link.Busy{{ID: "build-agent.server",
Container: "mesh-build-agent", State: "replaced-when-idle"}}}
if adoptionFingerprint(waiting) == adoptionFingerprint(quiet) {
t.Error("a container starting to wait is not news")
}
if adoptionFingerprint(still) != adoptionFingerprint(waiting) {
t.Error("a wait whose output changed reads as news on every ask")
}
if adoptionFingerprint(replaced) == adoptionFingerprint(waiting) {
t.Error("a container replaced after it waited is not news")
}
if !worthSaying(link.Report{Busy: replaced.Busy}) {
t.Error("a replacement after a wait is not worth saying")
}
}
// While a container waits, a pass is asked for every so often, not at the reconcile's five minutes; with
// nothing waiting, none (novox/hq ADR 0335 §2).
func TestAWaitIsAskedAgainSoonerThanTheReconcile(t *testing.T) {
if BusyAskEvery >= ReconcileEvery || BusyAskEvery > time.Minute {
t.Fatalf("a waiting container is asked every %s", BusyAskEvery)
}
var reconciles atomic.Int32
queue := &link.Queue{Reconcile: func(context.Context) (link.Report, bool) {
reconciles.Add(1)
return link.Report{}, false
}}
ctx, cancel := context.WithCancel(context.Background())
defer cancel()
go queue.Run(ctx)
busyWaiting.Store(false)
t.Cleanup(func() { busyWaiting.Store(false) })
go busyPasses(ctx, queue, 10*time.Millisecond)
time.Sleep(100 * time.Millisecond)
if n := reconciles.Load(); n != 0 {
t.Fatalf("%d passes were asked with nothing waiting", n)
}
busyWaiting.Store(true)
deadline := time.Now().Add(2 * time.Second)
for reconciles.Load() == 0 && time.Now().Before(deadline) {
time.Sleep(10 * time.Millisecond)
}
if reconciles.Load() == 0 {
t.Fatal("no pass was asked while a container waits")
}
}
cmd/mesh-host/filling.go
-74
View File
@@ -1,74 +0,0 @@
package main
import (
"context"
"crypto/ed25519"
"errors"
"path/filepath"
"github.com/novox/mesh-host/internal/apply"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/filling"
"github.com/novox/mesh-host/internal/store"
"github.com/novox/mesh-host/internal/system"
)
// serveFilling writes filled files on the provisioner client's request for as long as this engine serves (novox/hq
// ADR 0349 §4). A socket that cannot be served is said once, and this engine goes on applying.
func serveFilling(ctx context.Context, opts options, signer ed25519.PublicKey, say func(string)) {
if err := filling.Serve(ctx, filling.Socket, fillingEngine(opts, signer, say), say); err != nil {
say("filled files are not written on request on this machine: " + err.Error())
}
}
// fillingEngine is the write of a filled file as this engine does it: the declaration it keeps, verified, read under
// the apply's lock — the same two locks every apply takes, the process's and the one beside the state — and the file
// written through the apply (apply.WriteFilled), its outcome recorded in the node's state as an apply's is.
func fillingEngine(opts options, signer ed25519.PublicKey, say func(string)) *filling.Engine {
return &filling.Engine{
Declared: func() (*declaration.Declaration, error) {
raw, err := store.LoadDeclared(store.DeclaredPath(opts.state), signer)
if err != nil {
return nil, err
}
return declaration.Parse(raw)
},
UIDOf: filling.UIDOf,
Lock: func() (func(), error) {
applying.Lock()
unlock, err := store.Lock(opts.state, nil)
if err != nil {
applying.Unlock()
return nil, err
}
return func() { unlock(); applying.Unlock() }, nil
},
Place: func(ctx context.Context, d *declaration.Declaration, path string, content []byte) (string, string, error) {
built, err := system.For(builtFor)
if err != nil {
return "", "", err
}
known, err := store.Load(opts.state)
if err != nil {
return "", "", err
}
report, updated, writeErr := apply.WriteFilled(ctx, built, d, known, path, content, apply.ExecRunner,
announceOr(say), apply.KeepIn(filepath.Dir(opts.state)), apply.WindowsIn(filepath.Dir(opts.state)))
// Saved whichever way it went, as an apply's: what was written is on the machine either way.
if err := store.Save(opts.state, updated); err != nil {
return "", "", errors.Join(writeErr, errors.New("written, and the node's state could not be saved: "+
err.Error()))
}
action, digest := "", ""
for _, o := range report.Outcomes {
if o.Target == path && o.Type == string(declaration.TypeFile) {
action = o.Action
}
}
if rec, ok := updated.At(string(declaration.TypeFile), path); ok {
digest = rec.Wrote
}
return action, digest, writeErr
},
}
}
cmd/mesh-host/filling_test.go
-105
View File
@@ -1,105 +0,0 @@
package main
import (
"context"
"crypto/ed25519"
"encoding/json"
"os"
"os/user"
"path/filepath"
"testing"
"time"
"github.com/novox/mesh-host/internal/filling"
"github.com/novox/mesh-host/internal/store"
)
// Defends novox/hq ADR 0349 §4 (issue 668): a write of a filled file and an apply are one at a time — the write
// takes the lock the apply holds, and writes only once the apply let it go — and the write is recorded in the
// node's state as an apply's outcome is.
func TestAWriteAndAnApplyOfTheSameFileSerialise(t *testing.T) {
was := builtFor
builtFor = "arch"
t.Cleanup(func() { builtFor = was })
me, err := user.Current()
if err != nil {
t.Fatal(err)
}
if me.Uid == "0" {
t.Skip("run as root, whose account is never the provisioner's")
}
dir := t.TempDir()
opts := options{state: filepath.Join(dir, "state.json")}
public, private, err := ed25519.GenerateKey(nil)
if err != nil {
t.Fatal(err)
}
path := filepath.Join(dir, "m.conf")
raw, err := json.Marshal(map[string]any{"declaration": 1, "provisioner_account": me.Username,
"filled_files": []string{path}, "resources": []any{map[string]any{"id": "m.conf", "type": "file",
"path": path, "content": "from the declaration\n", "owner-at-runtime": "owned", "owner": me.Username}}})
if err != nil {
t.Fatal(err)
}
if err := store.SaveDeclared(store.DeclaredPath(opts.state),
store.Declared{Declaration: raw, Signature: ed25519.Sign(private, raw)}); err != nil {
t.Fatal(err)
}
e := fillingEngine(opts, public, func(string) {})
uid := uint32(os.Getuid())
if err := e.Accepts(uid); err != nil {
t.Fatal(err)
}
// An apply holds the lock.
applying.Lock()
done := make(chan filling.Answer)
go func() {
done <- e.Write(context.Background(), uid, filling.Request{Path: path, Content: []byte("filled\n")})
}()
select {
case a := <-done:
applying.Unlock()
t.Fatalf("written while an apply held the lock: %+v", a)
case <-time.After(200 * time.Millisecond):
}
if _, err := os.Stat(path); err == nil {
applying.Unlock()
t.Fatal("the file was written while an apply held the lock")
}
applying.Unlock()
a := <-done
if a.Refused != "" || a.Failed != "" || a.Action != "created" {
t.Fatalf("the write after the apply: %+v", a)
}
got, err := os.ReadFile(path)
if err != nil || string(got) != "filled\n" {
t.Fatalf("the file holds %q (%v)", got, err)
}
known, err := store.Load(opts.state)
if err != nil {
t.Fatal(err)
}
if rec, ok := known.Find("m.conf"); !ok || rec.Wrote != a.Digest || rec.Target != path {
t.Errorf("the write is not recorded in the node's state: %+v, answered %q", rec, a.Digest)
}
// And the cross-process lock beside the state too: a write waits for a `reconcile` run by hand.
unlock, err := store.Lock(opts.state, nil)
if err != nil {
t.Fatal(err)
}
go func() {
done <- e.Write(context.Background(), uid, filling.Request{Path: path, Content: []byte("again\n")})
}()
select {
case a := <-done:
unlock()
t.Fatalf("written while another process held the apply's lock: %+v", a)
case <-time.After(200 * time.Millisecond):
}
unlock()
if a := <-done; a.Action != "updated" {
t.Fatalf("the second write: %+v", a)
}
}
cmd/mesh-host/local_network_test.go
-29
View File
@@ -1,29 +0,0 @@
package main
import (
"testing"
"github.com/novox/mesh-host/internal/link"
"github.com/novox/mesh-host/internal/outward"
"github.com/novox/mesh-host/internal/profile"
)
func reportOn(networks ...outward.Network) link.Report {
return link.Report{Outward: []string{"eth0"},
Profile: profileAsReported(profile.Profile{Machine: &profile.Machine{Networks: networks}})}
}
// **A change of the networks a machine is attached to is said unasked** (novox/hq ADR 0340 §7): a port
// reached from the local network is admitted from them, so a laptop that joins another network is said at
// the next reconcile, not at the next push — and the same networks again are not news.
func TestAChangedNetworkIsSaidUnasked(t *testing.T) {
w := &adoptionWatch{}
home := reportOn(outward.Network{Link: "eth0", Network: "192.168.1.0/24"})
w.said(home)
if w.differs(reportOn(outward.Network{Link: "eth0", Network: "192.168.1.0/24"})) {
t.Fatal("the same networks read as a change, so the machine would speak on every reconcile")
}
if !w.differs(reportOn(outward.Network{Link: "eth0", Network: "10.20.0.0/16"})) {
t.Fatal("a machine that moved to another network does not say so until it is pushed to")
}
}
cmd/mesh-host/main.go
+7 -94
View File
@@ -24,7 +24,6 @@ import (
"sort"
"strings"
"sync"
"sync/atomic"
"syscall"
"text/tabwriter"
"time"
@@ -1130,7 +1129,7 @@ func runLink(ctx context.Context, opts options) error {
// The one liveness judge on this machine (novox/hq ADR 0240), made before anything applies: every
// report this host makes carries what it says.
if j, err := liveness.Open(filepath.Join(filepath.Dir(opts.state), liveness.FileName),
&liveness.Exec{Run: apply.ExecRunner, System: builtFor}); j != nil {
&liveness.Exec{Run: apply.ExecRunner}); j != nil {
if err != nil {
say(err.Error())
}
@@ -1242,16 +1241,6 @@ func runLink(ctx context.Context, opts options) error {
Say: say,
Timeout: opts.timeout,
}
// And the move of a value sealed to this machine into the vault, asked by the controller (design 57's V3,
// issue 676): the vault is asked over this queue's link, so the queue is named once it exists.
queue.MoveSecret = moveSecretFrom(opts.state, queue)
// And this machine's sealing key stated to the vault in its own name, at start and whenever it changes (ADR 0372,
// issue 786), whether or not any value was ever moved in.
go keepNodeKeyStated(ctx, func() (identity.SealingKey, error) {
return identity.LoadSealingKey(identity.SealingKeyPath(opts.state))
}, func(verb string, args any) (json.RawMessage, error) {
return queue.AskVault(ctx, verb, args)
}, say, nodeKeyRetryFrom, nodeKeyRetryUpTo)
worker := make(chan struct{})
go func() {
defer close(worker)
@@ -1270,7 +1259,6 @@ func runLink(ctx context.Context, opts options) error {
queue.ReconcileDue()
}
go holdTheMachine(ctx, queue)
go busyPasses(ctx, queue, BusyAskEvery)
// And whether what it runs stays up, looked at on its own clock and said when it changes (novox/hq
// ADR 0240) — between applies, which is when a container crash-loops.
if judging != nil {
@@ -1290,9 +1278,6 @@ func runLink(ctx context.Context, opts options) error {
// And mesh-cli is answered on this machine: the account from the kernel, the line asked of the controller
// (novox/hq ADR 0272).
go serveMeshCLI(aside, queue, say)
// And a filled file is written on the provisioner client's request, one at a time with the apply (novox/hq ADR
// 0349 §4).
go serveFilling(aside, opts, mine.Membership.Signer, say)
held := link.HoldRoused(aside, membership, queue, say, opts.timeout, rousedBySignal(ctx))
// The act in hand finishes and is kept before this process exits: an apply that stood aside with
@@ -1412,18 +1397,6 @@ func adoptionFingerprint(r link.Report) string {
// it could only speak when a declaration arrived, and a declaration cannot be composed until it
// has spoken: a machine waiting for a push that is waiting for the machine.
parts := []string{"firewall=" + r.Firewall, "outward=" + strings.Join(r.Outward, ",")}
// **And the networks those links are on** (novox/hq ADR 0340 §7): a port reached from the local
// network is admitted from them, so a laptop that joins another network, or a lease that moves, is
// said at the next reconcile rather than at the next push or the next day.
if networks := attachedNetworksOf(r); networks != "" {
parts = append(parts, "networks="+networks)
}
// And a container waiting because it said busy, or replaced after it waited (novox/hq ADR 0335): the walk
// waiting on this machine starts its gate at the report that says it replaced. Not its output, which
// may change on every ask.
for _, b := range r.Busy {
parts = append(parts, "busy "+b.ID+"="+b.State)
}
for _, h := range r.Held {
parts = append(parts, "held "+h.ID+"="+h.Changed)
}
@@ -1455,26 +1428,6 @@ func adoptionFingerprint(r link.Report) string {
return strings.Join(parts, "\n")
}
// attachedNetworksOf is the networks a report's profile says the machine is attached to, as one line: the
// part of what the machine is made of that the filter is composed from (novox/hq ADR 0340).
func attachedNetworksOf(r link.Report) string {
machine, ok := r.Profile["machine"].(map[string]any)
if !ok {
return ""
}
networks, ok := machine["networks"].([]any)
if !ok {
return ""
}
said := make([]string, 0, len(networks))
for _, n := range networks {
if entry, ok := n.(map[string]any); ok {
said = append(said, fmt.Sprintf("%v %v", entry["link"], entry["network"]))
}
}
return strings.Join(said, ",")
}
// applyState is how an apply went, as the fingerprint keeps it: refused, clean, or failed with what it did not
// apply and which keys its failure is said under.
func applyState(r link.Report) string {
@@ -1660,7 +1613,7 @@ func declaredResolvConf(d *declaration.Declaration) (string, string, bool) {
}
for _, r := range d.Resources {
f, ok := r.(*declaration.File)
if !ok || f.Path != network.ResolvConf || f.Seeded() || f.Into != "" {
if !ok || f.Path != network.ResolvConf || f.CreateOnce || f.Into != "" {
continue
}
module, ok := liveness.ModuleOf(f.ID)
@@ -1782,13 +1735,8 @@ func judgeWhatRuns(ctx context.Context, j *liveness.Judge, queue *link.Queue, sa
func healthAsReported(st liveness.Statement, ns *network.Statement) *link.Health {
// LibraryContract: this engine reads a resource's declared `health` and judges it (ADR 0240 Phase B), judges a
// user's declared `root` (novox/hq ADR 0266), and reads a declaration's identities, groups and libraries
// (novox/hq ADR 0328), a container's busy (ADR 0335) and a file's owner-at-runtime and the filled files (ADR
// 0349 §5) and the provisioner client's account (ADR 0349 §4), which is what tells the controller it may be sent
// any of those fields; and it writes a filled file only on the client's word (HandedContract, design 57 P7),
// which is what tells it it may hand a moved module's files to the provisioner here; and it opens the vault's box
// for a core credential (VaultBoxContract, ADR 0349 §16), which is what tells it it may send that box here; and it states its sealing key to the vault (NodeKeyContract, ADR
// 0372, issue 786).
h := &link.Health{Contract: link.NodeKeyContract, At: st.At.UTC(), Resources: []link.ResourceHealth{}}
// (novox/hq ADR 0328), which is what tells the controller it may be sent any of those fields.
h := &link.Health{Contract: link.OwnGroupContract, At: st.At.UTC(), Resources: []link.ResourceHealth{}}
for _, r := range st.Resources {
h.Resources = append(h.Resources, link.ResourceHealth{Module: r.Module, Resource: r.ID, Kind: r.Kind,
Target: r.Target, State: r.State, Reason: r.Reason, Since: r.Since.UTC(), Streak: r.Streak,
@@ -1812,8 +1760,7 @@ func withUnits(h *link.Health, us *units.Statement) *link.Health {
if us == nil || us.State == "" {
return h
}
h.Units = &link.UnitsHealth{State: us.State, Failed: []link.FailedUnit{}, Unread: us.Unread,
Away: us.Away}
h.Units = &link.UnitsHealth{State: us.State, Failed: []link.FailedUnit{}, Unread: us.Unread}
for _, f := range us.Failed {
if f.Module == "" {
h.Units.Failed = append(h.Units.Failed, link.FailedUnit{Unit: f.Unit, Scope: f.Scope, Load: f.Load,
@@ -1921,32 +1868,6 @@ func holdTheMachine(ctx context.Context, queue *link.Queue) {
}
}
// busyWaiting is whether the last apply left a container running because it said busy (novox/hq ADR 0335).
var busyWaiting atomic.Bool
// BusyAskEvery is how often a container waiting because it said busy is asked again (novox/hq ADR 0335 §2):
// each ask is a pass of its own, so a build that ends is followed by its replacement within this, not
// within the reconcile's five minutes — and the bound of 15 minutes is met within this of its end.
const BusyAskEvery = 20 * time.Second
// busyPasses asks for a pass every so often while a container waits because it said busy. **It asks; it
// does not apply**, as holdTheMachine: the queue's worker does, when its turn comes, so a pass never holds
// the machine for the wait, and a delivery waiting is that pass's apply.
func busyPasses(ctx context.Context, queue *link.Queue, every time.Duration) {
ticker := time.NewTicker(every)
defer ticker.Stop()
for {
select {
case <-ctx.Done():
return
case <-ticker.C:
if busyWaiting.Load() {
queue.ReconcileDue()
}
}
}
}
// reconcileKept is the reconcile's act, as the queue's worker runs it: hold the machine to what it
// was last told, read once it is this act's turn, and say what did not go.
//
@@ -1998,7 +1919,7 @@ func worthSaying(report link.Report) bool {
}
return len(report.Held) > 0 || report.Firewall != "" || len(report.Outward) > 0 ||
len(report.Filters) > 0 || report.FoundFirewall != nil || len(report.Windows) > 0 ||
len(report.Rollbacks) > 0 || len(report.Busy) > 0
len(report.Rollbacks) > 0
}
// announceOr is what the apply writes its detail with, given what the caller has to say things with.
@@ -2247,14 +2168,6 @@ func applyAndKeepHeld(ctx context.Context, opts options, raw []byte, signed *sto
for _, w := range outcome.Windows {
report.Windows = append(report.Windows, link.Window{Step: w.Step, Holds: w.Holds, Since: w.Opened, Until: w.Until})
}
// And a container left running because it said busy, or replaced after it waited (novox/hq ADR 0335):
// the walk waiting on this machine reads it. While one waits, this machine asks it again sooner than its
// reconcile (busyPasses).
for _, b := range outcome.Busy {
report.Busy = append(report.Busy, link.Busy{ID: b.ID, Container: b.Container, State: b.State,
Since: b.Since, Until: b.Until, Said: b.Said})
}
busyWaiting.Store(outcome.Waiting())
// And what runs here that nobody asked for (novox/hq ADR 0163).
if strays, err := apply.Strays(ctx, apply.ExecRunner, updated); err != nil {
fmt.Fprintf(os.Stderr, "mesh-host: applied, and could not list what else runs here: %v\n", err)
@@ -2367,7 +2280,7 @@ func sealOpener(statePath string) apply.Unseal {
if err != nil {
return nil, err
}
return key.Open(sealed)
return key.Unseal(sealed)
}
}
cmd/mesh-host/movesecret.go
-131
View File
@@ -1,131 +0,0 @@
package main
import (
"context"
"crypto/rand"
"crypto/subtle"
"encoding/base64"
"encoding/json"
"errors"
"fmt"
"os"
"time"
"golang.org/x/crypto/nacl/box"
"github.com/novox/mesh-host/internal/identity"
"github.com/novox/mesh-host/internal/link"
)
// The move of a value sealed to this machine into the vault (design 57's V3, issue 676; internal/link
// movesecret.go). The controller cannot open what it holds; this machine can open what was sealed to it, and so
// hands it to the vault, sealed to a key the vault made for the call. The value is in this process for that one
// call and written nowhere.
// vaultAsk asks one of the vault's verbs in this engine's own name and answers its result.
type vaultAsk func(verb string, args any) (json.RawMessage, error)
// moveSecretFrom is the engine's move: this machine's sealing key, read for each ask, and the vault asked over
// the link open now.
func moveSecretFrom(statePath string, queue *link.Queue) link.MoveSecretFunc {
return func(m link.MoveSecret) (link.MovedSecret, error) {
key, err := identity.LoadSealingKey(identity.SealingKeyPath(statePath))
if err != nil {
return link.MovedSecret{}, err
}
ask := func(verb string, args any) (json.RawMessage, error) {
return queue.AskVault(context.Background(), verb, args)
}
moved, err := moveIntoTheVault(ask, key, m)
if err != nil {
fmt.Fprintf(os.Stderr, "mesh-host: a value was not moved into the vault: %v\n", err)
}
return moved, err
}
}
// moveIntoTheVault opens the row's value with this machine's sealing key, hands it to the vault's move-in in its two
// steps, and answers what the vault kept once its read-back — sealed to a key made here for this one call — opens
// to the same value.
func moveIntoTheVault(ask vaultAsk, key identity.SealingKey, m link.MoveSecret) (link.MovedSecret, error) {
if m.SealedTo != "" && m.SealedTo != key.Public {
return link.MovedSecret{}, errors.New("the controller's store says this value was sealed to a sealing key this " +
"machine no longer has, so it cannot be opened here")
}
value, err := key.Unseal(m.Sealed)
if err != nil {
return link.MovedSecret{}, fmt.Errorf("the value does not open with this machine's sealing key: %w", err)
}
defer wipe(value)
// This machine's sealing key, stated in its own name: the key the vault seals the core's credentials to for the
// controller to carry here (novox/hq ADR 0349 §16, issue 763), never a key the controller names.
first := map[string]any{"origin": m.Origin, "generation": m.Generation, "sealing_key": key.Public}
if !m.MadeAt.IsZero() {
first["made_at"] = m.MadeAt.UTC().Format(time.RFC3339Nano)
}
if m.Own() {
first["node"], first["module"], first["name"] = m.Node, m.Module, m.Name
} else {
first["provider"], first["provision"] = m.Provider, m.Provision
first["consumer"], first["consumer_module"] = m.Consumer, m.ConsumerModule
if m.Local != "" {
first["local"] = m.Local
}
}
raw, err := ask("move-in", first)
if err != nil {
return link.MovedSecret{}, fmt.Errorf("the vault did not open the move: %w", err)
}
var opened struct {
Call string `json:"call"`
SealTo string `json:"seal_to"`
Secret string `json:"secret"`
}
if err := json.Unmarshal(raw, &opened); err != nil || opened.Call == "" || opened.SealTo == "" {
return link.MovedSecret{}, errors.New("the vault answered no call and no key to seal to")
}
sealed, err := identity.Seal(opened.SealTo, value)
if err != nil {
return link.MovedSecret{}, fmt.Errorf("the vault's key for the call is not a sealing key: %w", err)
}
backPub, backPriv, err := box.GenerateKey(rand.Reader)
if err != nil {
return link.MovedSecret{}, err
}
defer wipe(backPriv[:])
raw, err = ask("move-in", map[string]any{"call": opened.Call, "sealed": sealed,
"read_back_to": base64.StdEncoding.EncodeToString(backPub[:])})
if err != nil {
return link.MovedSecret{}, fmt.Errorf("the vault did not take %s: %w", opened.Secret, err)
}
var kept struct {
Name string `json:"name"`
Generation int `json:"generation"`
Unchanged bool `json:"unchanged"`
ReadBack string `json:"read_back"`
}
if err := json.Unmarshal(raw, &kept); err != nil || kept.Name == "" || kept.ReadBack == "" {
return link.MovedSecret{}, fmt.Errorf("the vault answered %s with no read-back", opened.Secret)
}
blob, err := base64.StdEncoding.DecodeString(kept.ReadBack)
if err != nil {
return link.MovedSecret{}, fmt.Errorf("the vault's read-back of %s is not sealed", kept.Name)
}
back, ok := box.OpenAnonymous(nil, blob, backPub, backPriv)
if !ok {
return link.MovedSecret{}, fmt.Errorf("the vault's read-back of %s does not open with the key made for it", kept.Name)
}
defer wipe(back)
if subtle.ConstantTimeCompare(back, value) != 1 {
return link.MovedSecret{}, fmt.Errorf("the vault keeps %s, but it read back another value than the one handed "+
"in: the move of it did not hold", kept.Name)
}
return link.MovedSecret{Secret: kept.Name, Generation: kept.Generation, Unchanged: kept.Unchanged}, nil
}
func wipe(b []byte) {
for i := range b {
b[i] = 0
}
}
cmd/mesh-host/movesecret_test.go
-134
View File
@@ -1,134 +0,0 @@
package main
import (
"crypto/rand"
"encoding/base64"
"encoding/json"
"errors"
"strings"
"testing"
"golang.org/x/crypto/nacl/box"
"github.com/novox/mesh-host/internal/identity"
"github.com/novox/mesh-host/internal/link"
)
// standInVault answers move-in's two steps as the vault does: a key made for the call, the value opened with it,
// and the value kept read back sealed to the key the second step sent. lie makes its read-back another value.
type standInVault struct {
t *testing.T
pub *[32]byte
priv *[32]byte
opened map[string]any
kept string
lie string
refuse string
carried []string
}
func (v *standInVault) ask(verb string, args any) (json.RawMessage, error) {
raw, _ := json.Marshal(args)
v.carried = append(v.carried, string(raw))
if verb != "move-in" {
v.t.Fatalf("asked %s", verb)
}
if v.refuse != "" {
return nil, errors.New(v.refuse)
}
var a map[string]any
_ = json.Unmarshal(raw, &a)
if a["call"] == nil {
v.opened = a
pub, priv, _ := box.GenerateKey(rand.Reader)
v.pub, v.priv = pub, priv
return json.Marshal(map[string]string{"call": "c1", "seal_to": base64.StdEncoding.EncodeToString(pub[:]),
"secret": "own/anchor/postgres/superuser-password"})
}
blob, _ := base64.StdEncoding.DecodeString(a["sealed"].(string))
value, ok := box.OpenAnonymous(nil, blob, v.pub, v.priv)
if !ok {
v.t.Fatal("what the engine sealed does not open with the call's key")
}
v.kept = string(value)
back := v.kept
if v.lie != "" {
back = v.lie
}
key, _ := base64.StdEncoding.DecodeString(a["read_back_to"].(string))
var to [32]byte
copy(to[:], key)
boxed, _ := box.SealAnonymous(nil, []byte(back), &to, rand.Reader)
return json.Marshal(map[string]any{"name": "own/anchor/postgres/superuser-password", "generation": 1,
"read_back": base64.StdEncoding.EncodeToString(boxed)})
}
const theValue = "the-superusers-password"
func aMoveOf(t *testing.T, key identity.SealingKey) link.MoveSecret {
t.Helper()
sealed, err := identity.Seal(key.Public, []byte(theValue))
if err != nil {
t.Fatal(err)
}
return link.MoveSecret{Node: "anchor", By: "the move", MoveSecretAsk: link.MoveSecretAsk{Module: "postgres",
Name: "superuser-password", Origin: "made", Generation: "g-1", Sealed: sealed, SealedTo: key.Public}}
}
// The engine opens what was sealed to its machine, hands it to the vault sealed to the call's key — never in the
// clear — and answers what the vault kept only once the read-back opens to the same value.
func TestTheEngineMovesAValueSealedToItAndChecksTheReadBack(t *testing.T) {
key, err := identity.GenerateSealingKey()
if err != nil {
t.Fatal(err)
}
vault := &standInVault{t: t}
moved, err := moveIntoTheVault(vault.ask, key, aMoveOf(t, key))
if err != nil {
t.Fatal(err)
}
if moved.Secret != "own/anchor/postgres/superuser-password" || moved.Generation != 1 || vault.kept != theValue {
t.Fatalf("moved %+v, the vault kept %q", moved, vault.kept)
}
if vault.opened["module"] != "postgres" || vault.opened["name"] != "superuser-password" || vault.opened["node"] != "anchor" ||
vault.opened["origin"] != "made" || vault.opened["generation"] != "g-1" ||
vault.opened["sealing_key"] != key.Public {
t.Errorf("the first step named %v", vault.opened)
}
if strings.Contains(strings.Join(vault.carried, " "), theValue) {
t.Fatal("the value crossed the bus in the clear")
}
}
// A read-back that is not the value handed in is a failed move, said; so is a value this machine cannot open, and
// the vault's refusal.
func TestTheEngineSaysAMoveThatDidNotHold(t *testing.T) {
key, _ := identity.GenerateSealingKey()
vault := &standInVault{t: t, lie: "something-else"}
if _, err := moveIntoTheVault(vault.ask, key, aMoveOf(t, key)); err == nil ||
!strings.Contains(err.Error(), "read back") {
t.Fatalf("a read-back of another value: %v", err)
}
other, _ := identity.GenerateSealingKey()
if _, err := moveIntoTheVault((&standInVault{t: t}).ask, key, aMoveOf(t, other)); err == nil ||
!strings.Contains(err.Error(), "sealing key") {
t.Fatalf("a value sealed to another key: %v", err)
}
refusing := &standInVault{t: t, refuse: "the move never replaces a value the vault holds"}
if _, err := moveIntoTheVault(refusing.ask, key, aMoveOf(t, key)); err == nil ||
!strings.Contains(err.Error(), "never replaces") {
t.Fatalf("the vault's refusal: %v", err)
}
// A pair credential names both its ends.
pair := aMoveOf(t, key)
pair.Module, pair.Name = "", ""
pair.Provider, pair.Provision, pair.Consumer, pair.ConsumerModule = "anchor", "database", "shanks", "gitea"
ends := &standInVault{t: t}
if _, err := moveIntoTheVault(ends.ask, key, pair); err != nil {
t.Fatal(err)
}
if ends.opened["provider"] != "anchor" || ends.opened["consumer"] != "shanks" || ends.opened["consumer_module"] != "gitea" ||
ends.opened["node"] != nil {
t.Errorf("the pair's first step named %v", ends.opened)
}
}
cmd/mesh-host/nodekey.go
-62
View File
@@ -1,62 +0,0 @@
package main
import (
"context"
"fmt"
"time"
"github.com/novox/mesh-host/internal/identity"
)
// This machine's sealing key, stated to the vault in the engine's own name (novox/hq ADR 0372, issue 786). The vault
// records it from the machine's own account alone and, when it is a first or a changed key, says the machine's core
// secrets changed so the controller asks for them again sealed to it. Without this a machine that never moved a
// secret in (adopted after feature 57) has no key at the vault. Nothing but the public half crosses.
const (
// nodeKeyRetryFrom and nodeKeyRetryUpTo bound the wait between tries while the vault does not answer: it starts
// after the bus and may not serve the verb yet.
nodeKeyRetryFrom = 5 * time.Second
nodeKeyRetryUpTo = 5 * time.Minute
// nodeKeyLook is how often the key file is read to see whether it changed.
nodeKeyLook = 30 * time.Second
)
// keepNodeKeyStated states the sealing key load answers to the vault at start, retrying with a growing wait from
// retryFrom to retryUpTo while it does not answer, and states it again whenever the key changes. It says one line
// when a key is stated and one when a try fails for another reason than the last, and returns when ctx ends.
func keepNodeKeyStated(ctx context.Context, load func() (identity.SealingKey, error), ask vaultAsk, say func(string),
retryFrom, retryUpTo time.Duration) {
stated, lastFailure := "", ""
wait := retryFrom
fail := func(why string) {
if why != lastFailure {
say("mesh-host: the sealing key was not stated to the vault, trying again: " + why)
lastFailure = why
}
wait = min(wait*2, retryUpTo)
}
look := min(nodeKeyLook, retryFrom)
for {
delay := look
key, err := load()
switch {
case err != nil:
fail(err.Error())
delay = wait
case key.Public != stated:
if _, err := ask("node-key", map[string]string{"key": key.Public}); err != nil {
fail(err.Error())
delay = wait
} else {
stated, lastFailure, wait = key.Public, "", retryFrom
say(fmt.Sprintf("mesh-host: this machine's sealing key %s is stated to the vault", key.Public))
}
}
select {
case <-ctx.Done():
return
case <-time.After(delay):
}
}
}
cmd/mesh-host/nodekey_test.go
-112
View File
@@ -1,112 +0,0 @@
package main
import (
"context"
"encoding/json"
"errors"
"sync"
"testing"
"time"
"github.com/novox/mesh-host/internal/identity"
"github.com/novox/mesh-host/internal/link"
)
// Issue 786 (novox/hq ADR 0372): the engine states its own sealing key to the vault's node-key at start, retries
// while the vault does not answer, and states it again when the key changes; the same key is not stated twice.
type keyVault struct {
mu sync.Mutex
stated []string
fail int
verbs []string
}
func (v *keyVault) ask(verb string, args any) (json.RawMessage, error) {
v.mu.Lock()
defer v.mu.Unlock()
v.verbs = append(v.verbs, verb)
if v.fail > 0 {
v.fail--
return nil, errors.New("the vault does not answer")
}
raw, _ := json.Marshal(args)
var a struct {
Key string `json:"key"`
}
_ = json.Unmarshal(raw, &a)
v.stated = append(v.stated, a.Key)
return json.RawMessage(`{"recorded":true}`), nil
}
func (v *keyVault) said() []string {
v.mu.Lock()
defer v.mu.Unlock()
return append([]string(nil), v.stated...)
}
func waitFor(t *testing.T, what string, ok func() bool) {
t.Helper()
for i := 0; i < 400; i++ {
if ok() {
return
}
time.Sleep(5 * time.Millisecond)
}
t.Fatalf("timed out: %s", what)
}
func TestTheEngineStatesItsSealingKeyToTheVaultAtStartRetriesAndRestatesAChange(t *testing.T) {
first, _ := identity.GenerateSealingKey()
second, _ := identity.GenerateSealingKey()
var mu sync.Mutex
current := first
load := func() (identity.SealingKey, error) {
mu.Lock()
defer mu.Unlock()
return current, nil
}
vault := &keyVault{fail: 2}
var lines []string
var lmu sync.Mutex
say := func(l string) { lmu.Lock(); lines = append(lines, l); lmu.Unlock() }
ctx, cancel := context.WithCancel(context.Background())
defer cancel()
go keepNodeKeyStated(ctx, load, vault.ask, say, time.Millisecond, 5*time.Millisecond)
waitFor(t, "the first key stated after two refusals", func() bool { return len(vault.said()) == 1 })
if vault.said()[0] != first.Public {
t.Errorf("stated %q, want the engine's own key %q", vault.said()[0], first.Public)
}
vault.mu.Lock()
verbs := append([]string(nil), vault.verbs...)
vault.mu.Unlock()
for _, v := range verbs {
if v != "node-key" {
t.Errorf("asked %s", v)
}
}
// The same key again is not stated again.
time.Sleep(60 * time.Millisecond)
if n := len(vault.said()); n != 1 {
t.Errorf("the same key was stated %d times", n)
}
mu.Lock()
current = second
mu.Unlock()
waitFor(t, "the changed key stated", func() bool { return len(vault.said()) == 2 })
if vault.said()[1] != second.Public {
t.Errorf("stated %q after the change, want %q", vault.said()[1], second.Public)
}
lmu.Lock()
defer lmu.Unlock()
if len(lines) == 0 {
t.Error("the engine said nothing")
}
}
// The statement a host makes names this contract (issue 786), one above the vault's box.
func TestTheNodeKeyContractIsThirteen(t *testing.T) {
if link.NodeKeyContract != 13 {
t.Errorf("NodeKeyContract %d", link.NodeKeyContract)
}
}
cmd/mesh-host/units_test.go
+4 -7
View File
@@ -87,13 +87,10 @@ func TestTheStatementSaysAnAccountJudgedForRoot(t *testing.T) {
Groups: []string{"openrazer"}}, State: accounts.Healthy, Since: at},
}}
h := withAccounts(healthAsReported(liveness.Statement{At: at}, nil), as)
// At RootContract or later: an engine that states its sealing key to the vault says NodeKeyContract (ADR 0372, issue 786), one above
// VaultBoxContract (it opens the vault's box for the core's credentials, ADR 0349 §16, issue 755), one above HandedContract (design 57 P7), one above ProvisionerContract (novox/hq ADR
// 0349 §4), one above ADR 0349 §5's OwnershipContract.
if h.Contract < link.RootContract || link.RootContract != 3 || h.Contract != link.NodeKeyContract ||
link.NodeKeyContract != link.VaultBoxContract+1 || link.VaultBoxContract != link.HandedContract+1 ||
link.HandedContract != link.ProvisionerContract+1 || link.ProvisionerContract != link.OwnershipContract+1 || link.OwnershipContract != link.BusyContract+1 ||
link.BusyContract != link.OwnGroupContract+1 {
// At RootContract or later: an engine that also reads an identity's runner says OwnGroupContract (novox/hq ADR
// 0334), one above ADR 0328's LibraryContract.
if h.Contract < link.RootContract || link.RootContract != 3 || h.Contract != link.OwnGroupContract ||
link.OwnGroupContract != link.LibraryContract+1 {
t.Fatalf("contract %d", h.Contract)
}
got := map[string]string{}
cmd/mesh-host/vaultbox_nats_test.go
-154
View File
@@ -1,154 +0,0 @@
package main
// Issue 755's proof on a real bus (novox/hq ADR 0349 §16): a core credential the vault sealed to this machine crosses
// the bus inside the declaration, which the controller carries and cannot open, and this engine opens it with its
// sealing key and writes the value. A listener on every subject keeps every byte that crossed; the value is never
// among them. Run against the throwaway bus merge-check.sh starts (MESH_TEST_NATS).
import (
"context"
"encoding/base64"
"encoding/json"
"os"
"path/filepath"
"strings"
"sync"
"testing"
"time"
"git.novox.be/novox/mesh-sdk/go/seal"
"github.com/nats-io/nats.go"
"github.com/novox/mesh-host/internal/apply"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/identity"
"github.com/novox/mesh-host/internal/link"
"github.com/novox/mesh-host/internal/store"
"github.com/novox/mesh-host/internal/system"
)
const coreCredential = "the-provisioner-clients-bus-password-2025"
func TestNatsTheEngineFillsACoreCredentialFromTheVaultsBoxWithNoValueInTheClear(t *testing.T) {
url := os.Getenv("MESH_TEST_NATS")
if url == "" {
t.Skip("MESH_TEST_NATS unset")
}
// This machine's sealing key, where the engine reads it.
state := filepath.Join(t.TempDir(), "state.json")
key, err := identity.GenerateSealingKey()
if err != nil {
t.Fatal(err)
}
if err := os.WriteFile(identity.SealingKeyPath(state), []byte(key.Private), 0o600); err != nil {
t.Fatal(err)
}
// Every byte that crosses.
tap, err := nats.Connect(url)
if err != nil {
t.Fatal(err)
}
defer tap.Close()
var mu sync.Mutex
var crossed []string
if _, err := tap.Subscribe(">", func(m *nats.Msg) {
mu.Lock()
crossed = append(crossed, m.Subject+" "+string(m.Data))
mu.Unlock()
}); err != nil {
t.Fatal(err)
}
_ = tap.Flush()
// The engine hears its declaration on its own subject.
const node = "vaultbox"
engine, err := nats.Connect(url)
if err != nil {
t.Fatal(err)
}
defer engine.Close()
heard, err := engine.SubscribeSync(link.DeclareSubject(node))
if err != nil {
t.Fatal(err)
}
_ = engine.Flush()
// The vault's answer to the controller: the value in the ADR 0206 box, sealed to this machine's key. The
// controller carries it in the declaration as it is.
pub, err := key.PublicPEM()
if err != nil {
t.Fatal(err)
}
box, err := seal.Seal(coreCredential, pub)
if err != nil {
t.Fatal(err)
}
path := filepath.Join(t.TempDir(), "broker")
body, _ := json.Marshal(map[string]any{"declaration": 1, "resources": []map[string]any{
{"id": "need-provisioner-broker", "type": "file", "path": path, "vault-sealed": box}}})
controller, err := nats.Connect(url)
if err != nil {
t.Fatal(err)
}
defer controller.Close()
if err := controller.Publish(link.DeclareSubject(node), body); err != nil {
t.Fatal(err)
}
_ = controller.Flush()
msg, err := heard.NextMsg(5 * time.Second)
if err != nil {
t.Fatalf("the declaration did not reach the engine: %v", err)
}
d, err := declaration.Parse(msg.Data)
if err != nil {
t.Fatal(err)
}
sys, err := system.For("arch")
if err != nil {
t.Fatal(err)
}
if _, _, err := apply.Apply(context.Background(), sys, d, store.State{}, store.OriginCarried,
func(context.Context, string, ...string) (string, error) { return "", nil }, nil, sealOpener(state)); err != nil {
t.Fatal(err)
}
on, err := os.ReadFile(path)
if err != nil || string(on) != coreCredential {
t.Fatalf("the engine wrote %q, %v", on, err)
}
_ = tap.Flush()
time.Sleep(100 * time.Millisecond)
mu.Lock()
all := strings.Join(crossed, "\n")
mu.Unlock()
if !strings.Contains(all, link.DeclareSubject(node)) || !strings.Contains(all, "vault-sealed") {
t.Fatalf("the listener heard nothing of the declaration:\n%s", all)
}
for _, form := range []string{coreCredential, base64.StdEncoding.EncodeToString([]byte(coreCredential))} {
if strings.Contains(all, form) {
t.Fatal("the core credential crossed the bus in the clear")
}
}
}
// The engine opens a value the controller sealed (the NaCl box) as before, and the vault's box, with one key.
func TestTheEnginesOpenerOpensBothBoxes(t *testing.T) {
state := filepath.Join(t.TempDir(), "state.json")
key, _ := identity.GenerateSealingKey()
if err := os.WriteFile(identity.SealingKeyPath(state), []byte(key.Private), 0o600); err != nil {
t.Fatal(err)
}
open := sealOpener(state)
nacl, _ := identity.Seal(key.Public, []byte("old"))
if got, err := open(nacl); err != nil || string(got) != "old" {
t.Errorf("the controller's box opened to %q, %v", got, err)
}
pub, _ := key.PublicPEM()
box, _ := seal.Seal("new", pub)
raw, _ := json.Marshal(box)
if got, err := open(string(raw)); err != nil || string(got) != "new" {
t.Errorf("the vault's box opened to %q, %v", got, err)
}
}
examples/bundle_test.go
+1 -1
View File
@@ -49,7 +49,7 @@ func TestTheBundleCarriesTheFoundationAndTheControlPlaneAndNothingElse(t *testin
}
sort.Strings(images)
want := []string{"mesh-controller", "nats", "pgvector"}
want := []string{"mesh-controller", "nats", "postgres"}
if strings.Join(images, ",") != strings.Join(want, ",") {
t.Fatalf("the bundle carries %v; expected exactly %v.\n\n"+
"Adding one is a change to what every first node becomes, and to ADR 0006's "+
examples/foundation-first-node.lock
+19 -34
View File
@@ -28,12 +28,11 @@
// image, and this file is applied on a machine with no mesh to ask about anything (ADR 0006). The
// registry prefix is the lab's; the digests are upstream's.
//
// **The store is raised where the `postgres` module keeps it** (novox/hq ADR 0338): the container
// `postgres`, on the module's pgvector image, its data in the directory /var/lib/mesh-store owned by
// the image's postgres user, its superuser password read from the module's own-secret file — so the
// store step adopts it as it runs. Pinned upstream by the module's own digest, not through the lab's
// registry prefix, because the two must be the same reference (mesh-host internal/bootstrap holds
// them to each other against the module's manifest).
// The store's data is a NAMED VOLUME, not a directory on the machine. A directory the host
// creates is owned by root, and the database runs as somebody else inside the container — so it
// could not write, and the container crash-looped. A named volume lets the image set up its own
// ownership, and outlives the container, which is what you want for the thing holding the mesh's
// state.
{
"declaration": 1,
@@ -84,28 +83,17 @@
"boot": "enabled",
"restart-on": ["base-filter"]
},
// The store's data is a directory owned by the image's postgres user (999), made before the
// container starts: a directory the runtime made would be root's, and the database could not
// write it. Measured on the module's image 2026-10-13: it initialises a new store on a directory
// made so (novox/hq ADR 0338 §5).
{
"id": "store-data",
"type": "directory",
"path": "/var/lib/mesh-store",
"mode": "0700",
"owner": "999:70"
},
{
"id": "store",
"type": "container",
"name": "postgres",
"image": "pgvector/pgvector@sha256:cf134a767f474095eeba57e0117be8e568e011a63f33fbf252f14c9b760f8e6f",
"name": "mesh-store",
"image": "192.0.2.250:5000/postgres@sha256:7abf537131b66ed5af448d90653abf1679b0c7e9a1f07efdd4c3108a401b259a",
"env": {
"POSTGRES_PASSWORD": "bootstrap",
"PGDATA": "/var/lib/postgresql/data/pgdata"
},
"ports": ["5432:5432"],
"volumes": ["/var/lib/mesh-store:/var/lib/postgresql/data"]
"volumes": ["mesh-store-data:/var/lib/postgresql/data"]
},
// Over TCP, not the socket. While the store initialises it runs a temporary server on the
// socket ONLY, then stops it and starts the real one — so a socket check passes, the action
@@ -114,21 +102,21 @@
{
"id": "store-ready",
"type": "action",
"in": "postgres",
"in": "mesh-store",
"command": ["sh", "-c", "for i in $(seq 1 180); do pg_isready -h 127.0.0.1 -U postgres >/dev/null 2>&1 && exit 0; sleep 1; done; echo 'the store did not answer within 180s; its own last words follow'; pg_isready -h 127.0.0.1 -U postgres; tail -n 20 /var/lib/postgresql/data/log/*.log 2>/dev/null; exit 1"],
"verify": ["pg_isready", "-h", "127.0.0.1", "-U", "postgres"]
},
{
"id": "inventory-database",
"type": "action",
"in": "postgres",
"in": "mesh-store",
"command": ["sh", "-c", "psql -U postgres -c 'CREATE DATABASE inventory'"],
"verify": ["sh", "-c", "psql -U postgres -lqt | cut -d'|' -f1 | grep -qw inventory"]
},
{
"id": "identity-database",
"type": "action",
"in": "postgres",
"in": "mesh-store",
"command": ["sh", "-c", "psql -U postgres -c 'CREATE DATABASE identity'"],
"verify": ["sh", "-c", "psql -U postgres -lqt | cut -d'|' -f1 | grep -qw identity"]
},
@@ -138,20 +126,20 @@
{
"id": "licences-database",
"type": "action",
"in": "postgres",
"in": "mesh-store",
"command": ["sh", "-c", "psql -U postgres -c 'CREATE DATABASE licences'"],
"verify": ["sh", "-c", "psql -U postgres -lqt | cut -d'|' -f1 | grep -qw licences"]
},
{
"id": "context-schemas",
"type": "action",
"command": ["docker", "run", "--rm", "--network", "container:postgres",
"command": ["docker", "run", "--rm", "--network", "container:mesh-store",
"-e", "MESH_STORE_INVENTORY=postgres://postgres:bootstrap@127.0.0.1:5432/inventory?sslmode=disable",
"-e", "MESH_STORE_IDENTITY=postgres://postgres:bootstrap@127.0.0.1:5432/identity?sslmode=disable",
"-e", "MESH_STORE_LICENCES=postgres://postgres:bootstrap@127.0.0.1:5432/licences?sslmode=disable",
"192.0.2.250:5000/mesh-controller@sha256:c67db38439ff0aee242b467486765467bb95801f52175fc5727cc4e437338ace",
"migrate"],
"verify": ["sh", "-c", "docker exec postgres psql -U postgres -d inventory -tAc \"select to_regclass('public.node')\" | grep -qx node && docker exec postgres psql -U postgres -d identity -tAc \"select to_regclass('public.signing_key')\" | grep -qx signing_key && docker exec postgres psql -U postgres -d licences -tAc \"select to_regclass('public.licence')\" | grep -qx licence"]
"verify": ["sh", "-c", "docker exec mesh-store psql -U postgres -d inventory -tAc \"select to_regclass('public.node')\" | grep -qx node && docker exec mesh-store psql -U postgres -d identity -tAc \"select to_regclass('public.signing_key')\" | grep -qx signing_key && docker exec mesh-store psql -U postgres -d licences -tAc \"select to_regclass('public.licence')\" | grep -qx licence"]
},
// **The bus is raised where the `nats` module keeps it** (novox/hq ADR 0333 §1): the container
// `mesh-broker-nats`, its configuration and user list in /var/lib/nats-module/conf, its streams in
@@ -221,20 +209,17 @@
"path": "/var/lib/nats-module/conf/accounts.conf",
"mode": "0600",
"create-once": true,
"content": "// The first user list, carried by the installer because at genesis there is no mesh to\n// compose one. A bootstrap credential, rotated with the store's and replaced by the\n// controller's own composition from its first start onward.\naccounts {\n MESH {\n jetstream: enabled\n users = [\n { user: \"controller\", password: \"$2a$10$AHqJgOifIVbU41KmATiMhuXFs8xa7Wl2HuN4UVBCXdN2jIQzjqApy\", permissions: {\n publish: { allow: [\"$JS.ACK.CONTROL.controller.>\", \"$JS.ACK.EVENTS.controller.>\", \"$JS.ACK.DEAD_LETTER_NOTICES.controller.>\", \"$JS.API.>\", \"$KV.mesh-controller_asked.>\", \"$KV.mesh-controller_assignments.>\", \"$KV.mesh-controller_calls.>\", \"$KV.mesh-controller_hand-acts.>\", \"$KV.mesh-controller_conditions.>\", \"$KV.mesh-controller_condition-history.>\", \"$KV.mesh-controller_declared-config.>\", \"$KV.mesh-controller_lease.>\", \"$KV.mesh-controller_predefined.>\", \"$KV.SEAT_MESH_BUILD_MACHINE_cancelled.>\", \"$KV.SEAT_NODE_BUILD_AGENT_cancelled.>\", \"_INBOX.enrol.>\", \"mesh.assignment.>\", \"mesh.mod.*.tool.>\", \"mesh.node.>\", \"mesh.seat.mesh-build-machine.accept.>\", \"mesh.seat.node-build-agent.accept.>\", \"mesh.seat.mesh-build-machine.tool.>\", \"mesh.seat.node-build-agent.tool.>\", \"mesh.seat.mesh-controller.event.applied\", \"mesh.seat.mesh-controller.event.built-before\", \"mesh.seat.mesh-controller.event.refused\", \"mesh.seat.mesh-controller.event.rolled-back\", \"mesh.seat.mesh-controller.event.checked\", \"mesh.seat.mesh-controller.event.plan-moved\", \"mesh.seat.mesh-controller.event.assignment-changed\", \"mesh.seat.mesh-controller.event.declared-config-changed\", \"mesh.seat.mesh-controller.event.predefined-changed\", \"mesh.seat.mesh-controller.event.condition-raised\", \"mesh.seat.mesh-controller.event.condition-changed\", \"mesh.seat.mesh-controller.event.condition-cleared\", \"mesh.seat.mesh-controller.event.doctor-heartbeat\", \"mesh.seat.mesh-controller.event.secret-replaced\", \"mesh.seat.mesh-controller.event.healer-acted\", \"$SRV.INFO\", \"mesh.seat.node-intrusion-prevention.tool.banned.*\", \"mesh.seat.node-backup.tool.backed-up.*\", \"mesh.seat.node-backup.tool.now.*\", \"mesh.seat.mesh-delivery.tool.stalled\", \"mesh.seat.mesh-delivery.tool.close\", \"mesh.seat.node-launcher.tool.secret.*\", \"mesh.seat.mesh-delivery.tool.release\", \"mesh.seat.mesh-delivery.tool.stop\", \"mesh.seat.mesh-delivery.tool.times\", \"mesh.seat.mesh-delivery.tool.groups\", \"mesh.seat.node-service-manager.tool.restart.*\", \"mesh.seat.operator-channel.tool.tell\", \"mesh.seat.mesh-controller.tool.plans\", \"mesh.events.dead.>\", \"mesh.again.>\"] }\n subscribe: { allow: [\"$JS.API.>\", \"$JS.EVENT.ADVISORY.CONSUMER.MAX_DELIVERIES.>\", \"$JS.EVENT.ADVISORY.CONSUMER.DELETED.>\", \"_DELIVER.controller\", \"_DELIVER.controller.>\", \"_INBOX.controller.>\", \"mesh.control.>\", \"mesh.mod.*.event.provisioner.failing\", \"mesh.mod.*.event.provisioner.recovered\", \"mesh.mod.*.event.provisioner.retirement\", \"mesh.mod.gitea.event.pull.merged\", \"mesh.mod.gitea.event.pull.updated\", \"mesh.mod.mesh-catalog.event.catching-up\", \"mesh.mod.mesh-catalog.event.upgraded\", \"mesh.seat.mesh-build-machine.event.built\", \"mesh.seat.node-build-agent.event.built\", \"mesh.seat.operator-channel.event.decided.mesh-controller\", \"mesh.seat.mesh-controller.tool.>\", \"$SRV.PING\", \"$SRV.INFO\", \"$SRV.PING.mesh-controller\", \"$SRV.PING.mesh-controller.>\", \"$SRV.INFO.mesh-controller\", \"$SRV.INFO.mesh-controller.>\", \"$SRV.STATS\", \"$SRV.STATS.mesh-controller\", \"$SRV.STATS.mesh-controller.>\"] }\n allow_responses: { max: 1, ttl: \"1m\" }\n } }\n ]\n }\n}\n"
"content": "// The first user list, carried by the installer because at genesis there is no mesh to\n// compose one. A bootstrap credential, rotated with the store's and replaced by the\n// controller's own composition from its first start onward.\naccounts {\n MESH {\n jetstream: enabled\n users = [\n { user: \"controller\", password: \"$2a$10$AHqJgOifIVbU41KmATiMhuXFs8xa7Wl2HuN4UVBCXdN2jIQzjqApy\", permissions: {\n publish: { allow: [\"$JS.ACK.CONTROL.controller.>\", \"$JS.ACK.EVENTS.controller.>\", \"$JS.ACK.DEAD_LETTER_NOTICES.controller.>\", \"$JS.API.>\", \"$KV.mesh-controller_asked.>\", \"$KV.mesh-controller_calls.>\", \"$KV.mesh-controller_hand-acts.>\", \"$KV.mesh-controller_conditions.>\", \"$KV.mesh-controller_condition-history.>\", \"$KV.mesh-controller_lease.>\", \"$KV.SEAT_MESH_BUILD_MACHINE_cancelled.>\", \"$KV.SEAT_NODE_BUILD_AGENT_cancelled.>\", \"_INBOX.enrol.>\", \"mesh.assignment.>\", \"mesh.mod.*.tool.>\", \"mesh.node.>\", \"mesh.seat.mesh-build-machine.accept.>\", \"mesh.seat.node-build-agent.accept.>\", \"mesh.seat.mesh-build-machine.tool.>\", \"mesh.seat.node-build-agent.tool.>\", \"mesh.seat.mesh-controller.event.applied\", \"mesh.seat.mesh-controller.event.built-before\", \"mesh.seat.mesh-controller.event.refused\", \"mesh.seat.mesh-controller.event.rolled-back\", \"mesh.seat.mesh-controller.event.checked\", \"mesh.seat.mesh-controller.event.plan-moved\", \"mesh.seat.mesh-controller.event.condition-raised\", \"mesh.seat.mesh-controller.event.condition-changed\", \"mesh.seat.mesh-controller.event.condition-cleared\", \"mesh.seat.mesh-controller.event.doctor-heartbeat\", \"mesh.seat.mesh-controller.event.secret-replaced\", \"mesh.seat.mesh-controller.event.healer-acted\", \"$SRV.INFO\", \"mesh.seat.node-intrusion-prevention.tool.banned.*\", \"mesh.seat.node-backup.tool.backed-up.*\", \"mesh.seat.node-backup.tool.now.*\", \"mesh.seat.mesh-delivery.tool.stalled\", \"mesh.seat.mesh-delivery.tool.close\", \"mesh.seat.node-launcher.tool.secret.*\", \"mesh.seat.mesh-delivery.tool.release\", \"mesh.seat.mesh-delivery.tool.stop\", \"mesh.seat.mesh-delivery.tool.times\", \"mesh.seat.mesh-delivery.tool.groups\", \"mesh.seat.node-service-manager.tool.restart.*\", \"mesh.seat.mesh-controller.tool.plans\", \"mesh.events.dead.>\", \"mesh.again.>\"] }\n subscribe: { allow: [\"$JS.API.>\", \"$JS.EVENT.ADVISORY.CONSUMER.MAX_DELIVERIES.>\", \"$JS.EVENT.ADVISORY.CONSUMER.DELETED.>\", \"_DELIVER.controller\", \"_DELIVER.controller.>\", \"_INBOX.controller.>\", \"mesh.control.>\", \"mesh.mod.*.event.provisioner.failing\", \"mesh.mod.*.event.provisioner.recovered\", \"mesh.mod.*.event.provisioner.retirement\", \"mesh.mod.gitea.event.pull.merged\", \"mesh.mod.gitea.event.pull.updated\", \"mesh.mod.mesh-catalog.event.catching-up\", \"mesh.mod.mesh-catalog.event.upgraded\", \"mesh.seat.mesh-build-machine.event.built\", \"mesh.seat.node-build-agent.event.built\", \"mesh.seat.operator-channel.event.decided.mesh-controller\", \"mesh.seat.mesh-controller.tool.>\", \"$SRV.PING\", \"$SRV.INFO\", \"$SRV.PING.mesh-controller\", \"$SRV.PING.mesh-controller.>\", \"$SRV.INFO.mesh-controller\", \"$SRV.INFO.mesh-controller.>\", \"$SRV.STATS\", \"$SRV.STATS.mesh-controller\", \"$SRV.STATS.mesh-controller.>\"] }\n allow_responses: { max: 1, ttl: \"1m\" }\n } }\n ]\n }\n}\n"
},
// The upstream server, until the broker step adopts it: the `nats` module's own image, built by the
// mesh, takes its place there once, on these same directories (ADR 0333 §3), as the bus's planned
// step (ADR 0236). The release is the one the module's image is built on (its NATS_BASE, 2.11.17),
// pinned upstream by that same digest: the adoption changes no server release, so it can be undone
// by putting the old image back and is said reversible. The published port is the
// mesh, takes its place there once, on these same directories (ADR 0333 §3). The published port is the
// node's bus port, 5671 unless given, which every enrolment token names; the installer gives the
// module the setting that maps its declared 4222 to it (ADR 0333 §2).
{
"id": "broker",
"type": "container",
"name": "mesh-broker-nats",
"image": "nats@sha256:e4bf19f15fd3218814a4e3c9e0064e1334bd8aa20d5984b9f1a0afd084f8cc00",
"image": "192.0.2.250:5000/nats@sha256:b83efabe3e7def1e0a4a31ec6e078999bb17c80363f881df35edc70fcb6bb927",
"ports": ["5671:4222", "127.0.0.1:8222:8222"],
"volumes": ["/var/lib/mesh-broker-nats:/data", "/var/lib/nats-module/conf:/etc/nats:ro", "/var/lib/mesh-broker-tls:/tls:ro"],
"args": ["-c", "/etc/nats/nats.conf", "-js"]
@@ -242,8 +227,8 @@
{
"id": "broker-ready",
"type": "action",
"command": ["sh", "-c", "for i in $(seq 1 60); do docker run --rm --network host --entrypoint sh nats@sha256:e4bf19f15fd3218814a4e3c9e0064e1334bd8aa20d5984b9f1a0afd084f8cc00 -c 'nc -z 127.0.0.1 5671' >/dev/null 2>&1 && exit 0; sleep 1; done; exit 1"],
"verify": ["sh", "-c", "docker run --rm --network host --entrypoint sh nats@sha256:e4bf19f15fd3218814a4e3c9e0064e1334bd8aa20d5984b9f1a0afd084f8cc00 -c 'nc -z 127.0.0.1 5671'"]
"command": ["sh", "-c", "for i in $(seq 1 60); do docker run --rm --network host --entrypoint sh 192.0.2.250:5000/nats@sha256:b83efabe3e7def1e0a4a31ec6e078999bb17c80363f881df35edc70fcb6bb927 -c 'nc -z 127.0.0.1 5671' >/dev/null 2>&1 && exit 0; sleep 1; done; exit 1"],
"verify": ["sh", "-c", "docker run --rm --network host --entrypoint sh 192.0.2.250:5000/nats@sha256:b83efabe3e7def1e0a4a31ec6e078999bb17c80363f881df35edc70fcb6bb927 -c 'nc -z 127.0.0.1 5671'"]
},
{
"id": "control-plane",
go.mod
+1 -1
View File
@@ -3,7 +3,7 @@ module github.com/novox/mesh-host
go 1.26.0
require (
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261015210729-d3e182503c19
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261010191001-33917f91ac3c
github.com/nats-io/nats.go v1.54.0
golang.org/x/crypto v0.57.0
golang.org/x/sys v0.48.0
go.sum
+6 -2
View File
@@ -1,5 +1,9 @@
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261015210729-d3e182503c19 h1:mAMt3KFmrHzDiBv4MLL94opWcG9y73ISMrBUyisapQE=
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261015210729-d3e182503c19/go.mod h1:GFuZUElBZ9A++mxgIKo97aXXo+kV0uJ/UkbhQPPIbrY=
git.novox.be/novox/mesh-sdk/go v0.1.11-0.20261009143344-f047d0a4a970 h1:9tFDQsgmI+4X7/BpZGXIr+HemPKE7YddYGqWV0lINAI=
git.novox.be/novox/mesh-sdk/go v0.1.11-0.20261009143344-f047d0a4a970/go.mod h1:GFuZUElBZ9A++mxgIKo97aXXo+kV0uJ/UkbhQPPIbrY=
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261010190644-622b6b4f556d h1:BR8uwv4BqyLaBwITuLylrd9DRUnu0D5DjJw0RuQSXbk=
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261010190644-622b6b4f556d/go.mod h1:GFuZUElBZ9A++mxgIKo97aXXo+kV0uJ/UkbhQPPIbrY=
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261010191001-33917f91ac3c h1:l5onJwoIeH8yE/PjVlEeoPyXBsHp2NQiWLllz2fwX7s=
git.novox.be/novox/mesh-sdk/go v0.1.14-0.20261010191001-33917f91ac3c/go.mod h1:GFuZUElBZ9A++mxgIKo97aXXo+kV0uJ/UkbhQPPIbrY=
github.com/klauspost/compress v1.20.0 h1:a3C1ke2ohxFymNlb2HWAHjDeKCI90scRskErZkR0ezA=
github.com/klauspost/compress v1.20.0/go.mod h1:LUdAzn7YLVvxLpc7y3V1m40wESHTgc1422pwwBSKYuI=
github.com/nats-io/nats.go v1.54.0 h1:vsXoOxjHp/GmPUN+EcI7uOf/uB+iAP+kEsAFNQN0yzA=
internal/accounts/accounts.go
+1 -1
View File
@@ -119,7 +119,7 @@ func secretsOf(d *declaration.Declaration, account string) []string {
var out []string
for _, r := range d.Resources {
f, ok := r.(*declaration.File)
if !ok || f.Path == "" || (!f.Secret() && len(f.Secrets) == 0) || f.Owner == account {
if !ok || f.Path == "" || (f.Sealed == "" && len(f.Secrets) == 0) || f.Owner == account {
continue
}
out = append(out, f.Path)
internal/apply/apply.go
+33 -115
View File
@@ -15,7 +15,6 @@ import (
"crypto/sha256"
"encoding/base64"
"encoding/hex"
"encoding/json"
"errors"
"fmt"
"os"
@@ -93,11 +92,6 @@ type Outcome struct {
// reads is, for a container, the digest of each file it was created reading, by path — so
// the next apply can say which one changed (novox/hq 04-ISSUES/103).
reads map[string]string
// busySince is, for a container left running because it said busy, when the first apply that would
// have replaced it ran (novox/hq ADR 0335): kept in its record, so the bound holds across passes and
// across a restart of the node-engine. busy is what the report says of it.
busySince *time.Time
busy *BusyWait
}
// Report is what an apply did, in the order it did it.
@@ -114,20 +108,6 @@ type Report struct {
// 224): a machine whose server is stopped at 03:31 because its collector is running is
// working, not broken, and the report is where that difference is said.
Windows []Window `json:"windows,omitempty"`
// Busy is every container this apply left running because it said busy, or replaced after it had
// waited (novox/hq ADR 0335): a machine whose walk waits for it says why, and how it ended.
Busy []BusyWait `json:"busy,omitempty"`
}
// Waiting says whether a container of this apply waits to be replaced because it is busy: what makes the
// node-engine ask again sooner than its reconcile (novox/hq ADR 0335 §2).
func (r Report) Waiting() bool {
for _, b := range r.Busy {
if b.State == BusyWaiting {
return true
}
}
return false
}
// Changed reports whether anything about the machine actually moved. An apply that changed
@@ -138,10 +118,7 @@ func (r Report) Changed() bool {
// whose account's manager is not running, which nothing moved either (novox/hq ADR 0177);
// held still is a container a maintenance window stopped, which the apply did not touch
// (novox/hq issue 224).
// waits busy is a container left running because it said busy (novox/hq ADR 0335), which nothing
// moved either.
if o.Action != "unchanged" && o.Action != "held" && o.Action != "waiting" && o.Action != heldStill &&
o.Action != waitsBusy {
if o.Action != "unchanged" && o.Action != "held" && o.Action != "waiting" && o.Action != heldStill {
return true
}
}
@@ -782,16 +759,6 @@ func ApplyMindingWindows(
}
}
// **A filled file is the provisioner client's to write** (novox/hq ADR 0349 §5, design 57 P7): one writer per
// file. The declaration says where it goes and how it is written, and only the write verb (WriteFilled)
// writes it, on the client's word. This apply — the controller's send and the five-minute re-apply alike —
// neither writes it nor records it, and nothing that reads it is moved for it; so only what the verb placed
// is recorded, and only that is removed as an orphan when the file leaves the declaration.
if f, isFile := resource.(*declaration.File); isFile && d.Fills(f.Path) {
report.Outcomes = append(report.Outcomes, leftToTheClient(f, &known, handedFiles, &report, log))
continue
}
// The private network takes over the tunnel it found, ahead of the service that replaces
// it (novox/hq ADR 0105): its configuration kept, its unit stopped and disabled, never
// flushed. A failure here fails the service too — the mesh's interface is not started on a
@@ -1010,7 +977,33 @@ func ApplyMindingWindows(
kept, keptMode, keptOwner = from.Kept, from.KeptMode, from.KeptOwner
}
// Only now. The record follows the fact, never leads it.
recordApplied(&known, origin, resource, outcome, kept, keptMode, keptOwner)
known.Record(store.Applied{
Origin: origin,
ID: resource.Identity(), Type: string(resource.Kind()),
Target: outcome.Target, AppliedAt: time.Now().UTC(),
Wrote: outcome.wrote,
Into: outcome.into,
Kept: kept,
KeptMode: keptMode,
KeptOwner: keptOwner,
Reads: outcome.reads,
Stateless: outcome.stateless,
Scope: outcome.scope,
User: outcome.user,
Found: outcome.found,
DeclaredUnit: outcome.declared,
Shell: outcome.shell,
Linger: outcome.linger,
Groups: outcome.groups,
Unpacked: outcome.unpacked,
AsFound: outcome.asFound,
Declared: outcome.asFoundDeclared,
FoundAt: outcome.asFoundAt,
Holds: holds(resource),
})
if outcome.found != nil {
known.DropFound(resource.Identity())
}
// Its module has been taken, and what was held for it is now the mesh's. A file written
// into, or a service whose lifecycle is the machine's, replaced nothing that was found, so
// its outcome says what the apply did, not that a cutover happened; a hold from when it was
@@ -1053,13 +1046,10 @@ func ApplyMindingWindows(
Action: "forgotten", Detail: detail})
log(fmt.Sprintf(" forgotten %s (%s): %s", from.ID, from.Target, detail))
}
if outcome.busy != nil {
report.Busy = append(report.Busy, *outcome.busy)
}
if outcome.Action == heldStill || outcome.Action == waitsBusy {
if outcome.Action == heldStill {
// Not changed: nothing about it moved, so nothing that restarts on it restarts.
log(fmt.Sprintf(" %s %s (%s): %s", outcome.Action, outcome.ID, outcome.Target, outcome.Detail))
} else if moved(outcome) {
} else if outcome.Action != "unchanged" {
changed[resource.Identity()] = true
// With the detail, when there is one: "updated app" says a container was replaced;
// which file made that happen is what somebody reading the log at the time needs
@@ -1152,47 +1142,6 @@ func ApplyMindingWindows(
return report, known, nil
}
// recordApplied records what applying a resource did, once it is done: the record follows the fact, never leads it.
// kept, keptMode and keptOwner are where the original of what it replaced was kept, and how it was found. What an
// earlier apply found its unit as goes with it.
func recordApplied(known *store.State, origin string, resource declaration.Resource, outcome Outcome,
kept, keptMode, keptOwner string) {
known.Record(store.Applied{
Origin: origin,
ID: resource.Identity(), Type: string(resource.Kind()),
Target: outcome.Target, AppliedAt: time.Now().UTC(),
Wrote: outcome.wrote,
Into: outcome.into,
Kept: kept,
KeptMode: keptMode,
KeptOwner: keptOwner,
Reads: outcome.reads,
BusySince: outcome.busySince,
Stateless: outcome.stateless,
Scope: outcome.scope,
User: outcome.user,
Found: outcome.found,
DeclaredUnit: outcome.declared,
Shell: outcome.shell,
Linger: outcome.linger,
Groups: outcome.groups,
Unpacked: outcome.unpacked,
AsFound: outcome.asFound,
Declared: outcome.asFoundDeclared,
FoundAt: outcome.asFoundAt,
Holds: holds(resource),
})
if outcome.found != nil {
known.DropFound(resource.Identity())
}
}
// moved reports whether applying a resource moved it, so what restarts or reloads on it does: anything but
// unchanged, and not a container held still or waiting for its busy one to say idle.
func moved(outcome Outcome) bool {
return outcome.Action != "unchanged" && outcome.Action != heldStill && outcome.Action != waitsBusy
}
// afterWhatTheyRead is the resources in declared order, except that **a service comes after every
// resource it names under restart-on or reload-on** (novox/hq issue 260).
//
@@ -1261,9 +1210,7 @@ func afterWhatTheyRead(resources []declaration.Resource) []declaration.Resource
const guardPrefix = declaration.AdoptionPrefix + "guard"
// Unseal opens a value the mesh sealed to this node. Nil when the node has no sealing key, which
// makes every sealed file an error rather than a silently skipped one. A value the vault sealed (a core
// credential, novox/hq ADR 0349 §16) is given as its box's JSON, which begins with `{`; one the controller sealed
// is base64, which never does (identity.SealingKey.Open).
// makes every sealed file an error rather than a silently skipped one.
type Unseal func(sealed string) ([]byte, error)
func applyOne(ctx context.Context, sys system.System, r declaration.Resource, run Runner,
@@ -1603,16 +1550,7 @@ func applyFile(r *declaration.File, previous store.Applied, unseal Unseal, keepF
return out, fmt.Errorf(
"%s is sealed to this node and this node has no sealing key", r.Path)
}
sealed := r.Sealed
if r.VaultSealed != nil {
// The vault's box, handed to the opener as its JSON (Unseal).
raw, err := json.Marshal(r.VaultSealed)
if err != nil {
return out, err
}
sealed = string(raw)
}
opened, err := unseal(sealed)
opened, err := unseal(r.Sealed)
if err != nil {
return out, fmt.Errorf("cannot open %s: %w", r.Path, err)
}
@@ -1653,7 +1591,7 @@ func applyFile(r *declaration.File, previous store.Applied, unseal Unseal, keepF
// A seed that is already there is left exactly as it is — whatever has grown in it since is
// not the mesh's to put back (novox/hq issue 035). What this host records is what it once
// wrote, so a later declaration that changes the seed is not mistaken for drift either.
if r.Seeded() && existed {
if r.CreateOnce && existed {
out.wrote = previous.Wrote
if out.wrote == "" {
out.wrote = digestOf(string(existing))
@@ -2704,12 +2642,6 @@ func applyContainer(ctx context.Context, r *declaration.Container, run Runner,
// (novox/hq 04-ISSUES/009). Recreating is how a container gets restart-on, which a service
// already has.
reasons := restartedBy(r.RestartOn, changed)
// **A replacement owed since a busy wait began is still owed** (novox/hq ADR 0335): a restart-on trigger
// counts as changed only in the pass it changed, so a container left running then would otherwise read
// unchanged at the next pass, never be replaced, and drop from the report as if it had been.
if owed := owedSinceBusy(r, previous); owed != "" {
reasons = append(reasons, owed)
}
// What this container was created reading, as last recorded. By its declared id first, and by
// its NAME when that id has no record: the bundle's `store` becomes the postgres module's
@@ -2760,8 +2692,6 @@ func applyContainer(ctx context.Context, r *declaration.Container, run Runner,
// stop, so a container this apply found stopped by a window is one whose window it can see.
if win, held := in.windows.holding(r.Name, time.Now()); held {
out.reads = wasReading
// A replacement owed since a busy wait stays owed through the window (novox/hq ADR 0335).
out.busySince = previous.BusySince
out.Action = heldStill
out.Detail = fmt.Sprintf("held still by the maintenance window of %s since %s; left as it is, "+
"and converged by the first apply after the window closes (novox/hq issue 224)",
@@ -2777,15 +2707,6 @@ func applyContainer(ctx context.Context, r *declaration.Container, run Runner,
out.Action = "unchanged"
return out, nil
case existed:
// **A container that says it is busy is asked before it is replaced** (novox/hq ADR 0335 §2): left
// running while it says busy, within the bound, and replaced on the first pass it says idle. Only a
// running one: a stopped container holds no work in hand.
if r.Busy != nil && before.Running {
if waits := askBeforeReplacing(ctx, r, run, cri, previous, &out); waits {
out.reads = wasReading
return out, nil
}
}
if _, err := run(ctx, cri, "rm", "-f", r.Name); err != nil {
return out, fmt.Errorf("replacing container %s: %w", r.Name, err)
}
@@ -2886,9 +2807,6 @@ func applyContainer(ctx context.Context, r *declaration.Container, run Runner,
default:
out.Detail = "replaced; a container's configuration is fixed when it is created"
}
if out.busy != nil {
out.Detail += "; " + out.busy.Words()
}
}
return out, nil
}
internal/apply/busy.go
-153
View File
@@ -1,153 +0,0 @@
package apply
import (
"context"
"fmt"
"strings"
"time"
"unicode/utf8"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
)
// A container that says it is busy is replaced once it is idle, within a bound (novox/hq ADR 0335).
//
// **Asked, never waited on.** The apply asks the container once, and while it says busy leaves it running
// as it is and goes on with everything else of the declaration: the machine's lock is held for one ask,
// never for the wait. The wait is carried between passes in the container's record (store.Applied's
// BusySince), so the next pass — the node-engine asks again sooner than its reconcile while anything waits
// — asks again, and so does a node-engine started again in between: the bound runs from the first apply
// that would have replaced the container, whoever applied it.
// waitsBusy is the action of a container left running because it said busy (novox/hq ADR 0335 §2). Its own
// word rather than "unchanged": its declaration moved, and the machine is not yet what was declared.
const waitsBusy = "waits-busy"
// What a report says of a busy container (novox/hq ADR 0335 §2, §3).
const (
// BusyWaiting is a container left running because it said busy, waiting to be replaced.
BusyWaiting = "waiting"
// BusyReplacedIdle is a container replaced on the first pass it said idle, after it had waited.
BusyReplacedIdle = "replaced-when-idle"
// BusyReplacedAtBound is a container still busy at the bound, replaced anyway.
BusyReplacedAtBound = "replaced-at-bound"
)
// BusyWait is one container a busy command held, as a report says it.
type BusyWait struct {
// ID is the resource's id (`<module>.<resource>`), Container its runtime name.
ID string `json:"id"`
Container string `json:"container"`
// State is BusyWaiting, BusyReplacedIdle or BusyReplacedAtBound.
State string `json:"state"`
// Since is when the first apply that would have replaced it ran; Until when the bound replaces it.
Since time.Time `json:"since"`
Until time.Time `json:"until"`
// Said is the last output its busy command gave.
Said string `json:"said,omitempty"`
}
// Words is the wait in a sentence, for an outcome's detail and the console.
func (b BusyWait) Words() string {
since := b.Since.UTC().Format(time.RFC3339)
said := ""
if b.Said != "" {
said = "; its busy command last said: " + b.Said
}
switch b.State {
case BusyReplacedAtBound:
return fmt.Sprintf("replaced while still busy: it said busy since %s and the bound of %s ran out "+
"(novox/hq ADR 0335 §3)%s", since, declaration.BusyBound, said)
case BusyReplacedIdle:
return fmt.Sprintf("replaced once it said idle, after waiting since %s (novox/hq ADR 0335 §2)%s", since, said)
}
return fmt.Sprintf("busy: left running and replaced on the first pass it says idle, or at %s at the latest; "+
"waiting since %s (novox/hq ADR 0335 §2)%s", b.Until.UTC().Format(time.RFC3339), since, said)
}
// owedSinceBusy is the reason a container still owes the replacement it was left running for, or empty: its
// record keeps when its busy wait began, under the same name.
func owedSinceBusy(r *declaration.Container, previous store.Applied) string {
if previous.BusySince == nil || previous.Target != r.Name {
return ""
}
return "the replacement owed since its busy wait began at " + previous.BusySince.UTC().Format(time.RFC3339)
}
// busyClock is the clock the bound is read on. A test moves it.
var busyClock = time.Now
// busySaidMost is how much of a busy command's output a report carries: its last lines, enough to say what
// it held.
const busySaidMost = 400
// askBeforeReplacing asks a running container that declares busy whether it may be replaced now, and says
// in out what came of it. True when it waits: out is then the whole outcome of this pass. False when the
// caller replaces it — idle, or past the bound — with out carrying the wait it ended, if there was one.
func askBeforeReplacing(ctx context.Context, r *declaration.Container, run Runner, cri string,
previous store.Applied, out *Outcome) bool {
now := busyClock().UTC()
busy, said := askBusy(ctx, r, run, cri)
var since time.Time
waited := previous.BusySince != nil && previous.Target == r.Name
if waited {
since = previous.BusySince.UTC()
// A wait kept from a clock that has since gone back reads as starting now, never as later: the bound
// holds from no later than this pass.
if since.After(now) {
since = now
}
}
wait := BusyWait{ID: r.ID, Container: r.Name, Since: since, Said: said}
switch {
case !busy && !waited:
// Idle at the first ask: replaced at once, as a container that declares nothing is.
return false
case !busy:
wait.State, wait.Until = BusyReplacedIdle, since.Add(declaration.BusyBound)
out.busy = &wait
return false
}
if !waited {
since = now
wait.Since = now
}
wait.Until = since.Add(declaration.BusyBound)
if !now.Before(wait.Until) {
wait.State = BusyReplacedAtBound
out.busy = &wait
return false
}
wait.State = BusyWaiting
out.Action, out.Detail = waitsBusy, wait.Words()
out.busySince, out.busy = &since, &wait
return true
}
// askBusy runs a container's busy command inside it, by its shell, as an exec health check is run: busy is
// an exit of 0 and nothing else. An ask that could not be made, or did not answer within its timeout, is
// not busy: the bound is what keeps a container whose command never answers from being held.
func askBusy(ctx context.Context, r *declaration.Container, run Runner, cri string) (bool, string) {
ask, cancel := context.WithTimeout(ctx, r.Busy.Within())
defer cancel()
output, err := run(ask, cri, "exec", r.Name, "/bin/sh", "-c", r.Busy.Command)
said := strings.TrimSpace(output)
if err != nil && said == "" {
said = err.Error()
}
return err == nil, lastOf(said, busySaidMost)
}
// lastOf is the end of a text, at most n bytes, on one line.
func lastOf(s string, n int) string {
s = strings.Join(strings.Fields(s), " ")
if len(s) <= n {
return s
}
from := len(s) - n
for from < len(s) && !utf8.RuneStart(s[from]) {
from++
}
return "…" + s[from:]
}
internal/apply/busy_test.go
-433
View File
@@ -1,433 +0,0 @@
package apply
import (
"context"
"errors"
"path/filepath"
"slices"
"strings"
"sync"
"testing"
"time"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
)
// A container that says it is busy is replaced once it is idle, within a bound (novox/hq ADR 0335). The
// machine of these tests runs two containers, a build agent that declares `busy` and a web server that
// does not; every call that replaces a container is counted.
type aBusyMachine struct {
mu sync.Mutex
// containers is each container's name and the spec it was created with; running whether it runs.
containers map[string]string
busy bool // what the agent's busy command answers
said string // what it prints
asked int
calls []string
}
func newBusyMachine() *aBusyMachine {
return &aBusyMachine{containers: map[string]string{}}
}
func (m *aBusyMachine) run(_ context.Context, _ string, args ...string) (string, error) {
m.mu.Lock()
defer m.mu.Unlock()
switch args[0] {
case "info", "image":
return "27.0\n", nil
case "container": // inspect
name := args[len(args)-1]
spec, there := m.containers[name]
if !there {
return "", errors.New("no such container")
}
return "true\t" + spec, nil
case "exec":
m.asked++
m.calls = append(m.calls, "ask "+args[1])
if strings.Join(args[2:], " ") != "/bin/sh -c mesh-builder busy" {
return "", errors.New("asked something else: " + strings.Join(args[2:], " "))
}
if m.busy {
return m.said + "\n", nil
}
return m.said + "\n", errors.New("exited 1")
case "rm":
name := args[len(args)-1]
m.calls = append(m.calls, "rm "+name)
delete(m.containers, name)
case "run":
name := args[slices.Index(args, "--name")+1]
m.calls = append(m.calls, "create "+name)
for i, a := range args {
if a == "--label" && strings.HasPrefix(args[i+1], specLabel+"=") {
m.containers[name] = strings.TrimPrefix(args[i+1], specLabel+"=")
}
}
}
return "", nil
}
func (m *aBusyMachine) mark() int {
m.mu.Lock()
defer m.mu.Unlock()
return len(m.calls)
}
// since is every call since mark that replaced a container, or asked one.
func (m *aBusyMachine) since(mark int) []string {
m.mu.Lock()
defer m.mu.Unlock()
return append([]string(nil), m.calls[mark:]...)
}
func (m *aBusyMachine) answer(busy bool, said string) {
m.mu.Lock()
defer m.mu.Unlock()
m.busy, m.said = busy, said
}
// twoContainers is the declaration at a version: v moves the agent's image tag through its env, w the web's.
func twoContainers(t *testing.T, agent, web string) *declaration.Declaration {
t.Helper()
return parse(t, `{"declaration":1,"resources":[
{"id":"build-agent.server","type":"container","name":"mesh-build-agent","image":"`+pinned+`",
"env":{"V":"`+agent+`"},"busy":{"command":"mesh-builder busy","timeout":"5s"}},
{"id":"web.server","type":"container","name":"web","image":"`+pinned+`","env":{"V":"`+web+`"}}
]}`)
}
// atClock moves the bound's clock for one test.
func atClock(t *testing.T, at *time.Time) {
t.Helper()
was := busyClock
busyClock = func() time.Time { return *at }
t.Cleanup(func() { busyClock = was })
}
func busyOf(r Report, id string) (BusyWait, bool) {
for _, b := range r.Busy {
if b.ID == id {
return b, true
}
}
return BusyWait{}, false
}
func applyBusy(t *testing.T, m *aBusyMachine, d *declaration.Declaration, known store.State) (Report, store.State) {
t.Helper()
report, state, err := Apply(context.Background(), archHost(t), d, known, store.OriginDeclared, m.run, nil, nil)
if err != nil {
t.Fatalf("the apply failed: %v", err)
}
return report, state
}
// The decision's first check: an apply that would replace a container whose busy command exits 0 leaves it
// running, applies the rest, and replaces it on the first pass the command exits non-zero.
func TestABusyContainerIsLeftRunningAndReplacedOnTheFirstPassItSaysIdle(t *testing.T) {
now := time.Date(2026, 10, 13, 1, 0, 0, 0, time.UTC)
atClock(t, &now)
m := newBusyMachine()
_, state := applyBusy(t, m, twoContainers(t, "1", "1"), store.State{})
// A merge moves both: the agent holds a build, the web server holds nothing.
m.answer(true, "building novox/mesh-controller (check of #280) since 00:58")
mark := m.mark()
report, state := applyBusy(t, m, twoContainers(t, "2", "2"), state)
if got := m.since(mark); !slices.Equal(got, []string{"ask mesh-build-agent", "rm web", "create web"}) {
t.Fatalf("a busy agent was not left running while the rest applied: %v", got)
}
o := outcomeOf(report, "build-agent.server")
if o.Action != waitsBusy {
t.Fatalf("a busy container was reported %q, want %q: %+v", o.Action, waitsBusy, o)
}
for _, says := range []string{"busy", "first pass it says idle", "check of #280", "ADR 0335"} {
if !strings.Contains(o.Detail, says) {
t.Errorf("the outcome does not say %q: %s", says, o.Detail)
}
}
if (Report{Outcomes: []Outcome{o}}).Changed() {
t.Errorf("a container left running reads as a change: %+v", o)
}
b, said := busyOf(report, "build-agent.server")
if !said || b.State != BusyWaiting || !b.Since.Equal(now) || !b.Until.Equal(now.Add(15*time.Minute)) ||
b.Container != "mesh-build-agent" || !strings.Contains(b.Said, "check of #280") {
t.Fatalf("the report does not say the agent waits, since when and until when: %+v", report.Busy)
}
if !report.Waiting() {
t.Error("a report with a container waiting does not say so")
}
if rec, _ := state.Find("build-agent.server"); rec.BusySince == nil || !rec.BusySince.Equal(now) {
t.Fatalf("the wait is not kept in the container's record: %+v", rec)
}
// Asked again a minute later, still busy: still left running, the wait counted from the first apply.
now = now.Add(time.Minute)
mark = m.mark()
report, state = applyBusy(t, m, twoContainers(t, "2", "2"), state)
if got := m.since(mark); !slices.Equal(got, []string{"ask mesh-build-agent"}) {
t.Fatalf("a second pass while busy did more than ask: %v", got)
}
if b, _ := busyOf(report, "build-agent.server"); b.State != BusyWaiting || !b.Since.Equal(now.Add(-time.Minute)) {
t.Fatalf("the second pass does not keep the wait from the first apply: %+v", report.Busy)
}
// The build ends: the next pass replaces it, and says it waited.
now = now.Add(time.Minute)
m.answer(false, "idle")
mark = m.mark()
report, state = applyBusy(t, m, twoContainers(t, "2", "2"), state)
if got := m.since(mark); !slices.Equal(got, []string{"ask mesh-build-agent", "rm mesh-build-agent", "create mesh-build-agent"}) {
t.Fatalf("the agent was not replaced on the first pass it said idle: %v", got)
}
if o := outcomeOf(report, "build-agent.server"); o.Action != "updated" || !strings.Contains(o.Detail, "once it said idle") {
t.Errorf("the replacement does not say it waited: %+v", o)
}
if b, _ := busyOf(report, "build-agent.server"); b.State != BusyReplacedIdle {
t.Errorf("the report does not say the agent was replaced when idle: %+v", report.Busy)
}
if report.Waiting() {
t.Error("a report with nothing waiting says something waits")
}
if rec, _ := state.Find("build-agent.server"); rec.BusySince != nil {
t.Errorf("the wait is still kept after the replacement: %+v", rec)
}
// And the pass after that has nothing to say about it.
report, _ = applyBusy(t, m, twoContainers(t, "2", "2"), state)
if len(report.Busy) != 0 {
t.Errorf("a container replaced is still said: %+v", report.Busy)
}
}
// An idle container is replaced at once, as one that declares nothing is: asked once, nothing waits.
func TestAnIdleContainerIsReplacedAtOnce(t *testing.T) {
m := newBusyMachine()
_, state := applyBusy(t, m, twoContainers(t, "1", "1"), store.State{})
m.answer(false, "")
mark := m.mark()
report, _ := applyBusy(t, m, twoContainers(t, "2", "1"), state)
if got := m.since(mark); !slices.Equal(got, []string{"ask mesh-build-agent", "rm mesh-build-agent", "create mesh-build-agent"}) {
t.Fatalf("an idle agent was not replaced at once: %v", got)
}
if len(report.Busy) != 0 {
t.Errorf("an idle container replaced at once is said as a wait: %+v", report.Busy)
}
// And a container whose declaration did not move is never asked.
mark = m.mark()
applyBusy(t, m, twoContainers(t, "2", "1"), state)
if got := m.since(mark); slices.Contains(got, "ask mesh-build-agent") {
t.Errorf("a container nothing replaces was asked: %v", got)
}
}
// Past the bound it is replaced anyway, the report carrying the command's last output — and the bound holds
// across a restart of the node-engine: the wait is read back from the state on disk.
func TestABusyContainerIsReplacedAtTheBoundEvenAfterARestart(t *testing.T) {
now := time.Date(2026, 10, 13, 1, 0, 0, 0, time.UTC)
atClock(t, &now)
m := newBusyMachine()
_, state := applyBusy(t, m, twoContainers(t, "1", "1"), store.State{})
m.answer(true, "building novox/mesh-host (a check that never ends)")
_, state = applyBusy(t, m, twoContainers(t, "2", "1"), state)
// The node-engine stops and starts again: what it knows is what it saved.
path := filepath.Join(t.TempDir(), "state.json")
if err := store.Save(path, state); err != nil {
t.Fatal(err)
}
reloaded, err := store.Load(path)
if err != nil {
t.Fatal(err)
}
now = now.Add(14 * time.Minute)
mark := m.mark()
report, reloaded := applyBusy(t, m, twoContainers(t, "2", "1"), reloaded)
if got := m.since(mark); slices.Contains(got, "rm mesh-build-agent") {
t.Fatalf("a container busy within the bound was replaced after a restart: %v", got)
}
if b, _ := busyOf(report, "build-agent.server"); b.State != BusyWaiting || !b.Since.Equal(now.Add(-14*time.Minute)) {
t.Fatalf("a restart started the wait again: %+v", report.Busy)
}
now = now.Add(time.Minute)
mark = m.mark()
report, _ = applyBusy(t, m, twoContainers(t, "2", "1"), reloaded)
if got := m.since(mark); !slices.Equal(got, []string{"ask mesh-build-agent", "rm mesh-build-agent", "create mesh-build-agent"}) {
t.Fatalf("a container still busy at the bound was not replaced: %v", got)
}
b, _ := busyOf(report, "build-agent.server")
if b.State != BusyReplacedAtBound || !strings.Contains(b.Said, "never ends") {
t.Fatalf("the report does not say it replaced a busy container, with its last output: %+v", report.Busy)
}
if o := outcomeOf(report, "build-agent.server"); !strings.Contains(o.Detail, "still busy") ||
!strings.Contains(o.Detail, "never ends") {
t.Errorf("the outcome does not say a busy container was replaced, with what it said: %s", o.Detail)
}
}
// While one container waits, a send that changes another module on the same machine is applied at once: the
// apply returns after one ask, and holds nothing for the wait.
func TestASendToAnotherModuleIsAppliedWhileOneWaits(t *testing.T) {
now := time.Date(2026, 10, 13, 1, 0, 0, 0, time.UTC)
atClock(t, &now)
m := newBusyMachine()
_, state := applyBusy(t, m, twoContainers(t, "1", "1"), store.State{})
m.answer(true, "building")
_, state = applyBusy(t, m, twoContainers(t, "2", "1"), state)
// The next send moves only the web server.
mark := m.mark()
done := make(chan struct{})
var report Report
go func() {
defer close(done)
report, _ = applyBusy(t, m, twoContainers(t, "2", "2"), state)
}()
select {
case <-done:
case <-time.After(5 * time.Second):
t.Fatal("an apply while a container waits did not return")
}
got := m.since(mark)
if !slices.Contains(got, "create web") || slices.Contains(got, "rm mesh-build-agent") {
t.Fatalf("the other module's send was not applied at once beside the wait: %v", got)
}
if o := outcomeOf(report, "web.server"); o.Action != "updated" {
t.Errorf("the web server was reported %q: %+v", o.Action, o)
}
}
// A container's busy is held to its shape where it arrives.
func TestBusyIsReadAndHeldToItsShape(t *testing.T) {
d := twoContainers(t, "1", "1")
c := d.Resources[0].(*declaration.Container)
if c.Busy == nil || c.Busy.Command != "mesh-builder busy" || c.Busy.Within() != 5*time.Second {
t.Fatalf("busy was not read: %+v", c.Busy)
}
for name, raw := range map[string]string{
"no command": `"busy":{"command":" "}`,
"a long timeout": `"busy":{"command":"x","timeout":"5m"}`,
"not a duration": `"busy":{"command":"x","timeout":"soon"}`,
"an unknown key": `"busy":{"command":"x","every":"1m"}`,
"on a step": `"run-once":true,"busy":{"command":"x"}`,
} {
_, err := declaration.Parse([]byte(`{"declaration":1,"resources":[{"id":"a.b","type":"container",` +
`"name":"b","image":"` + pinned + `",` + raw + `}]}`))
if err == nil {
t.Errorf("%s: a busy of that shape was accepted", name)
}
}
}
// A replacement owed to a restart-on trigger is still owed after the wait: a trigger that is no file or
// directory (a service, an identity's or a library's change) is in no container's spec and counts as changed only
// in the pass it changed, so the container must not read unchanged at the next pass and drop from the report as if
// replaced (novox/hq ADR 0335).
func TestAReplacementOwedToARestartTriggerIsMadeOnceIdle(t *testing.T) {
now := time.Date(2026, 10, 13, 1, 0, 0, 0, time.UTC)
atClock(t, &now)
d := parse(t, `{"declaration":1,"resources":[
{"id":"build-agent.server","type":"container","name":"mesh-build-agent","image":"`+pinned+`",
"restart-on":["identity:build-agent"],"busy":{"command":"mesh-builder busy"}}
]}`)
r := d.Resources[0].(*declaration.Container)
in := inputs{declares: map[string]string{}}
m := newBusyMachine()
m.containers["mesh-build-agent"] = containerSpec(r, in)
ctx := context.Background()
// The trigger moved this pass, and the agent holds a build: left running.
m.answer(true, "building")
out, err := applyContainer(ctx, r, m.run, map[string]bool{"identity:build-agent": true}, in, store.Applied{})
if err != nil || out.Action != waitsBusy || out.busySince == nil {
t.Fatalf("a busy container was not left running for its trigger: %+v %v", out, err)
}
record := store.Applied{ID: r.ID, Target: r.Name, BusySince: out.busySince}
// The next pass: nothing triggers now, and the replacement is still owed.
now = now.Add(time.Minute)
m.answer(false, "idle")
mark := m.mark()
out, err = applyContainer(ctx, r, m.run, map[string]bool{}, in, record)
if err != nil {
t.Fatal(err)
}
if got := m.since(mark); !slices.Equal(got, []string{"ask mesh-build-agent", "rm mesh-build-agent", "create mesh-build-agent"}) {
t.Fatalf("the replacement owed to the trigger was dropped: %v (%+v)", got, out)
}
if out.busy == nil || out.busy.State != BusyReplacedIdle || out.busySince != nil {
t.Fatalf("the replacement does not say it was made once idle: %+v", out)
}
}
// A wait kept from a clock that later went back is never longer than the bound from now.
func TestAWaitFromAClockThatWentBackIsClampedToNow(t *testing.T) {
now := time.Date(2026, 10, 13, 1, 0, 0, 0, time.UTC)
atClock(t, &now)
m := newBusyMachine()
_, state := applyBusy(t, m, twoContainers(t, "1", "1"), store.State{})
m.answer(true, "building")
_, state = applyBusy(t, m, twoContainers(t, "2", "1"), state)
now = now.Add(-time.Hour) // the clock went back an hour
report, _ := applyBusy(t, m, twoContainers(t, "2", "1"), state)
b, _ := busyOf(report, "build-agent.server")
if b.State != BusyWaiting || b.Since.After(now) || !b.Until.Equal(now.Add(15*time.Minute)) {
t.Fatalf("a wait from a later clock is not clamped to now: %+v (now %s)", b, now)
}
}
// An ask that fails — no shell in the image, a timeout, a paused container — is not busy: replaced at once.
func TestAnAskThatFailsIsIdle(t *testing.T) {
m := newBusyMachine()
_, state := applyBusy(t, m, twoContainers(t, "1", "1"), store.State{})
failing := func(ctx context.Context, name string, args ...string) (string, error) {
if args[0] == "exec" {
m.mu.Lock()
m.calls = append(m.calls, "ask "+args[1])
m.mu.Unlock()
return "", errors.New("OCI runtime exec failed: exec: \"/bin/sh\": no such file or directory")
}
return m.run(ctx, name, args...)
}
mark := m.mark()
report, _, err := Apply(context.Background(), archHost(t), twoContainers(t, "2", "1"), state,
store.OriginDeclared, failing, nil, nil)
if err != nil {
t.Fatal(err)
}
if got := m.since(mark); !slices.Equal(got, []string{"ask mesh-build-agent", "rm mesh-build-agent", "create mesh-build-agent"}) {
t.Fatalf("an ask that failed held the container: %v", got)
}
if len(report.Busy) != 0 {
t.Errorf("an ask that failed is said as a wait: %+v", report.Busy)
}
}
// A stopped container holds no work: it is replaced without being asked.
func TestAStoppedContainerIsNotAsked(t *testing.T) {
m := newBusyMachine()
_, state := applyBusy(t, m, twoContainers(t, "1", "1"), store.State{})
m.answer(true, "would say busy")
stopped := func(ctx context.Context, name string, args ...string) (string, error) {
if args[0] == "container" && args[len(args)-1] == "mesh-build-agent" {
out, err := m.run(ctx, name, args...)
return strings.Replace(out, "true\t", "false\t", 1), err
}
return m.run(ctx, name, args...)
}
mark := m.mark()
if _, _, err := Apply(context.Background(), archHost(t), twoContainers(t, "2", "1"), state,
store.OriginDeclared, stopped, nil, nil); err != nil && !strings.Contains(err.Error(), "not running") {
t.Fatal(err)
}
got := m.since(mark)
if slices.Contains(got, "ask mesh-build-agent") || !slices.Contains(got, "rm mesh-build-agent") {
t.Fatalf("a stopped container was asked, or not replaced: %v", got)
}
}
internal/apply/filled.go
-230
View File
@@ -1,230 +0,0 @@
package apply
import (
"context"
"fmt"
"slices"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
"github.com/novox/mesh-host/internal/system"
)
// WriteFilled writes what the provisioner's client filled into one file the declaration names as filled (novox/hq
// ADR 0349 §4), and makes what reads the file take the change, exactly as an apply that changed it would.
//
// **The declared file says everything but the content.** Its path, mode, owner and who owns it at runtime are the
// declaration's; the client's request carries the content alone. So the file is applied as the declared one with
// that content in place of the declaration's: an owned file whole through applyFile, a seed only when absent, a
// managed file's keys through applyInto or its block through applyBlock — the same guards on where nothing is
// placed, the same Keep of an original found there, the same record. What the declaration seals or fills with
// `${secret:}` the client has already filled, so the content is written as it came; a file that carried a secret
// keeps the mode a secret is written with when the declaration gives none.
//
// **Then what reads it, in the apply's order** (afterWhatTheyRead): a service naming it under restart-on is
// restarted and under reload-on reloaded, a process or container naming it under restart-on is restarted, and a
// running container that reads it at creation — an env-file, a file mounted into it — is recreated, because what
// it read is recorded by digest and this write changed the digest. Each of those is applied and recorded as an
// apply does, and what moves goes on moving what names it. Nothing else of the declaration is touched.
//
// The caller holds the apply's lock: a write and an apply never write one file at once. log is said the path, the
// action and the digest, never the content.
func WriteFilled(
ctx context.Context,
sys system.System,
d *declaration.Declaration,
known store.State,
path string,
content []byte,
run Runner,
log func(string),
keep Keep,
windows *Windows,
) (report Report, _ store.State, _ error) {
if log == nil {
log = func(string) {}
}
if !d.Fills(path) {
return report, known, &Error{Resource: path, Err: fmt.Errorf("this node's declaration does not name %s "+
"as a file the provisioner fills, so nothing was written", path)}
}
var declared []*declaration.File
for _, r := range d.Resources {
if f, ok := r.(*declaration.File); ok && f.Path == path {
declared = append(declared, f)
}
}
if len(declared) != 1 {
// The parser refuses a filled path no file declares; two at one path would need the request to say whose
// keys or block it carries, and it carries the content only.
return report, known, &Error{Resource: path, Err: fmt.Errorf("%d resources declare %s, and a filled file "+
"is written for exactly one, so nothing was written", len(declared), path)}
}
file := filledAs(declared[0], content)
if held, isHeld := known.HeldAt(file.ID); isHeld {
return report, known, &Error{Resource: file.ID, Err: fmt.Errorf("%s is held as found for a module not "+
"yet taken (%s), so nothing was written", path, held.Why)}
}
in := inputs{declares: map[string]string{}, known: &known, windows: windows, away: newStoreAway(windows, log),
groups: groupsWanted(d.Resources), handed: Handed{}}
refreshRuntimeRoots(ctx)
in.refused = refusedPlaces(d.Resources)
in.agentHomes = rootNeverHomes(d)
in.dirsBefore = map[string]bool{}
// A container that reads the file and mounts a library whose disk or mount is not there is held, as a full apply
// holds it: not created over the missing disk (novox/hq ADR 0328 §9).
in.held = heldLibraries(ctx, d.Libraries)
for _, r := range d.Resources {
in.declares[r.Identity()] = declaredDigest(r)
}
changed := map[string]bool{}
// The file, as the apply loop applies it.
was, _ := known.Find(file.ID)
var keepFound Keep
if (was.ID == "" || was.Target != file.Path) && !known.Recorded(string(declaration.TypeFile), file.Path) {
keepFound = keep
}
outcome, err := applyOne(ctx, sys, file, run, changed, in, was, nil, keepFound)
if err != nil {
log(fmt.Sprintf(" failed %s (%s): %v", file.ID, file.Path, err))
return report, known, &Error{Resource: file.ID, Err: err, Done: report}
}
recordKeeping(&known, file, outcome, was)
report.Outcomes = append(report.Outcomes, outcome)
if !moved(outcome) {
return report, known, nil
}
changed[file.ID] = true
line := fmt.Sprintf(" %s %s (%s), filled: %s", outcome.Action, file.ID, file.Path, outcome.wrote)
if outcome.Detail != "" {
line += ": " + outcome.Detail
}
log(line)
// What reads it, in the apply's order, each moving what names it in turn.
var failures []*Error
for _, r := range afterWhatTheyRead(d.Resources) {
if r.Identity() == file.ID || !readsChange(r, path, changed) {
continue
}
if _, isHeld := known.HeldAt(r.Identity()); isHeld {
log(fmt.Sprintf(" held %s (%s): held as found for a module not yet taken, so it was not restarted "+
"for %s", r.Identity(), r.Target(), file.ID))
continue
}
was, _ := known.Find(r.Identity())
o, err := applyOne(ctx, sys, r, run, changed, in, was, nil, nil)
if err != nil {
if o.found != nil {
known.KeepFound(r.Identity(), store.OriginDeclared, *o.found)
}
failures = append(failures, &Error{Resource: r.Identity(), Err: err})
log(fmt.Sprintf(" failed %s (%s): %v", r.Identity(), o.Target, err))
continue
}
if o.Action == "waiting" {
if o.found != nil {
known.KeepFound(r.Identity(), store.OriginDeclared, *o.found)
}
report.Outcomes = append(report.Outcomes, o)
log(fmt.Sprintf(" waiting %s (%s): %s", o.ID, o.Target, o.Detail))
continue
}
recordKeeping(&known, r, o, was)
report.Outcomes = append(report.Outcomes, o)
if o.busy != nil {
report.Busy = append(report.Busy, *o.busy)
}
if moved(o) {
changed[r.Identity()] = true
line := fmt.Sprintf(" %s %s (%s)", o.Action, o.ID, o.Target)
if o.Detail != "" {
line += ": " + o.Detail
}
log(line)
} else if o.Action != "unchanged" {
log(fmt.Sprintf(" %s %s (%s): %s", o.Action, o.ID, o.Target, o.Detail))
}
}
if len(failures) > 0 {
first := failures[0]
first.Done = report
first.Others = len(failures) - 1
return report, known, first
}
return report, known, nil
}
// filledAs is the declared file with the filled content in place of the declaration's.
func filledAs(declared *declaration.File, content []byte) *declaration.File {
f := *declared
f.Content, f.Sealed, f.VaultSealed, f.Bytes, f.Secrets = string(content), "", nil, "", nil
if f.Mode == "" && (declared.Secret() || len(declared.Secrets) > 0) {
// Written as the declaration would have written it: a file carrying a credential is not world-readable.
f.Mode = "0600"
}
return &f
}
// readsChange reports whether applying r is what makes it take a change of what changed: a service's restart-on or
// reload-on, a process's or container's restart-on, or a running container that reads path at creation.
func readsChange(r declaration.Resource, path string, changed map[string]bool) bool {
switch res := r.(type) {
case *declaration.Service:
return len(restartedBy(res.RestartOn, changed)) > 0 || len(restartedBy(res.ReloadOn, changed)) > 0
case *declaration.Process:
return len(restartedBy(res.RestartOn, changed)) > 0
case *declaration.Container:
if len(restartedBy(res.RestartOn, changed)) > 0 {
return true
}
if res.RunOnce || res.Schedule != "" {
return false
}
if slices.Contains(res.EnvFile, path) {
return true
}
for _, v := range res.Volumes {
if mountSource(v) == path {
return true
}
}
}
return false
}
// recordKeeping records r as applied, carrying where the original it replaced was kept from its record when this
// write kept none — for as long as the resource is recorded at the same target, as the apply loop carries it.
func recordKeeping(known *store.State, r declaration.Resource, o Outcome, was store.Applied) {
kept, keptMode, keptOwner := o.kept, o.keptMode, o.keptOwner
if kept == "" && was.Target == o.Target {
kept, keptMode, keptOwner = was.Kept, was.KeptMode, was.KeptOwner
}
recordApplied(known, store.OriginDeclared, r, o, kept, keptMode, keptOwner)
}
// leftToTheClient is a filled file as an apply leaves it (novox/hq ADR 0349 §5, design 57 P7): untouched, and said
// so. A file the mesh wrote at the same path under another id, going in this apply, is handed to it as an apply hands
// a whole file (novox/hq ADR 0223): its record — what the mesh wrote there, and the original it kept — is the filled
// file's now, so the verb judges what is there against it, and the file is neither removed nor written.
func leftToTheClient(f *declaration.File, known *store.State, handedFiles map[string]store.Applied, report *Report,
log func(string)) Outcome {
if from, handed := handedFiles[f.ID]; handed {
if _, recorded := known.Find(f.ID); !recorded {
rec := from
rec.ID = f.ID
known.Record(rec)
}
delete(handedFiles, f.ID)
known.Forget(from.ID)
detail := "no longer declared; " + f.ID + " declares the file now, filled by the provisioner's client, so " +
"it was handed to it, not removed"
report.Outcomes = append(report.Outcomes, Outcome{ID: from.ID, Type: from.Type, Target: from.Target,
Action: "forgotten", Detail: detail})
log(fmt.Sprintf(" forgotten %s (%s): %s", from.ID, from.Target, detail))
}
return Outcome{ID: f.ID, Type: string(declaration.TypeFile), Target: f.Path, Action: "unchanged",
Detail: "filled by the provisioner's client, which alone has it written"}
}
internal/apply/filled_apply_test.go
-218
View File
@@ -1,218 +0,0 @@
package apply
import (
"context"
"fmt"
"os"
"path/filepath"
"strings"
"testing"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
)
// Defends novox/hq ADR 0349 §5, one writer per file (design 57 P7, issue 673): a file the declaration names as
// filled is written only through the write verb (WriteFilled), on the provisioner client's word. An apply — the
// controller's send and the five-minute re-apply alike — never writes it, never records it, and removes it as an
// orphan only when the verb placed it.
// An apply leaves a filled file alone, whatever who owns it at runtime says: absent stays absent and is not
// recorded, and what the verb wrote stays as it wrote it through every re-apply.
func TestAnApplyNeverWritesAFilledFile(t *testing.T) {
for _, c := range []struct{ name, fields, content string }{
{"owned", `"content":"","mode":"0640","owner-at-runtime":"owned"`, "port = 8080\n"},
{"seed", `"content":"","mode":"0600","owner-at-runtime":"seed"`, "user default on\n"},
{"managed block", `"into":"block","content":"","owner-at-runtime":"managed"`, "10.42.0.1 ace\n"},
} {
t.Run(c.name, func(t *testing.T) {
path := filepath.Join(t.TempDir(), "the-file")
d := filledDecl(t, path, c.fields, "")
report, state, err := Apply(context.Background(), archHost(t), d, store.State{}, store.OriginDeclared,
noServices, nil, nil)
if err != nil {
t.Fatal(err)
}
if _, err := os.Stat(path); err == nil {
t.Fatalf("an apply wrote a filled file the client has not filled: %q", readText(t, path))
}
if _, ok := state.Find("m.conf"); ok {
t.Errorf("an apply recorded a filled file it did not write")
}
if o := outcomeOf(report, "m.conf"); o.Action != "unchanged" || !strings.Contains(o.Detail, "provisioner") {
t.Errorf("an apply did not say it left the file to the provisioner's client: %+v", o)
}
_, state, err = WriteFilled(context.Background(), archHost(t), d, state, path, []byte(c.content),
noServices, nil, nil, nil)
if err != nil {
t.Fatal(err)
}
placed, _ := state.Find("m.conf")
want, wantMode := readText(t, path), modeAt(t, path)
// The re-apply, twice over: nothing written, the verb's record kept as it wrote it.
for i := 0; i < 2; i++ {
var said []string
report, state, err = Apply(context.Background(), archHost(t), d, state, store.OriginDeclared,
noServices, func(line string) { said = append(said, line) }, nil)
if err != nil {
t.Fatal(err)
}
if got := readText(t, path); got != want || modeAt(t, path) != wantMode {
t.Fatalf("re-apply %d wrote the filled file again: %q, want %q", i+1, got, want)
}
if o := outcomeOf(report, "m.conf"); o.Action != "unchanged" {
t.Errorf("re-apply %d: %+v", i+1, o)
}
if rec, _ := state.Find("m.conf"); rec.Wrote != placed.Wrote || rec.Target != path {
t.Errorf("re-apply %d changed the verb's record: %+v, was %+v", i+1, rec, placed)
}
if len(said) != 0 {
t.Errorf("re-apply %d said something of an untouched file: %q", i+1, said)
}
}
})
}
}
// What reads a filled file is moved by the verb's write, once, never again by an apply: a service under
// restart-on is not restarted and a container reading it is not recreated by the re-apply.
func TestAnApplyMovesNothingForAFilledFile(t *testing.T) {
env := filepath.Join(t.TempDir(), "forge.env")
d := filledDecl(t, env, `"content":"","mode":"0600","owner-at-runtime":"owned"`,
`{"id":"forge.server","type":"container","name":"forge","image":"`+pinned+`","env-file":["`+env+`"]}`)
m := &machine{containers: map[string]*fakeContainer{}}
_, state, err := WriteFilled(context.Background(), archHost(t), d, store.State{}, env, []byte("PORT=5432\n"),
m.run, nil, nil, nil)
if err != nil {
t.Fatal(err)
}
m.asked = nil
if _, _, err := Apply(context.Background(), archHost(t), d, state, store.OriginDeclared, m.run, nil,
nil); err != nil {
t.Fatal(err)
}
if m.removed("forge") || m.did("docker run") {
t.Errorf("the re-apply recreated the container reading the filled file: %v", m.asked)
}
if got := readText(t, env); got != "PORT=5432\n" {
t.Errorf("the re-apply wrote the filled file: %q", got)
}
conf := filepath.Join(t.TempDir(), "daemon.conf")
d = filledDecl(t, conf, `"content":"","owner-at-runtime":"owned"`,
`{"id":"m.daemon","type":"service","unit":"m.service","state":"running","restart-on":["m.conf"]}`)
var commands []string
_, state, err = WriteFilled(context.Background(), archHost(t), d, store.State{}, conf, []byte("level = 1\n"),
recordingServices(&commands), nil, nil, nil)
if err != nil {
t.Fatal(err)
}
commands = nil
if _, _, err := Apply(context.Background(), archHost(t), d, state, store.OriginDeclared,
recordingServices(&commands), nil, nil); err != nil {
t.Fatal(err)
}
if n := count(commands, "systemctl stop m.service"); n != 0 {
t.Errorf("the re-apply restarted what reads the filled file: %v", commands)
}
}
// A filled file the verb never placed is not the mesh's to remove: one found there, or put there by hand, stays
// when the file leaves the declaration.
func TestAFilledFileTheVerbNeverPlacedIsNotRemoved(t *testing.T) {
path := filepath.Join(t.TempDir(), "the-file")
if err := os.WriteFile(path, []byte("the machine's own\n"), 0o644); err != nil {
t.Fatal(err)
}
d := filledDecl(t, path, `"content":"","owner-at-runtime":"owned"`, "")
_, state, err := Apply(context.Background(), archHost(t), d, store.State{}, store.OriginDeclared, noServices,
nil, nil)
if err != nil {
t.Fatal(err)
}
gone := elsewhere(t)
report, state, err := Apply(context.Background(), archHost(t), gone, state, store.OriginDeclared, noServices,
nil, nil)
if err != nil {
t.Fatal(err)
}
if got := readText(t, path); got != "the machine's own\n" {
t.Errorf("a filled file the verb never placed was touched when it left: %q", got)
}
if o := outcomeOf(report, "m.conf"); o.Action == "removed" {
t.Errorf("a filled file the verb never placed was removed: %+v", o)
}
if _, ok := state.Find("m.conf"); ok {
t.Errorf("a record of the filled file was left")
}
}
// What the verb placed is the mesh's: when the file leaves the declaration it is removed as an orphan, as a file
// an apply wrote is.
func TestAFilledFileTheVerbPlacedIsRemovedWhenItLeaves(t *testing.T) {
path := filepath.Join(t.TempDir(), "the-file")
d := filledDecl(t, path, `"content":"","owner-at-runtime":"owned"`, "")
_, state, err := Apply(context.Background(), archHost(t), d, store.State{}, store.OriginDeclared, noServices,
nil, nil)
if err != nil {
t.Fatal(err)
}
_, state, err = WriteFilled(context.Background(), archHost(t), d, state, path, []byte("port = 8080\n"),
noServices, nil, nil, nil)
if err != nil {
t.Fatal(err)
}
gone := elsewhere(t)
report, state, err := Apply(context.Background(), archHost(t), gone, state, store.OriginDeclared, noServices,
nil, nil)
if err != nil {
t.Fatal(err)
}
if _, err := os.Stat(path); err == nil {
t.Errorf("a filled file the verb placed stayed when it left the declaration: %q", readText(t, path))
}
if o := outcomeOf(report, "m.conf"); o.Action != "removed" {
t.Errorf("not removed: %+v", o)
}
if _, ok := state.Find("m.conf"); ok {
t.Errorf("its record stayed")
}
}
// A file the mesh wrote the old way and that is now filled keeps its record for the verb to judge against, and is
// not written over by the apply that hands it to the client; a file at the same path recorded under another id
// is handed to the filled one, not removed.
func TestAFileThatMovesToTheClientIsHandedNotRewritten(t *testing.T) {
path := filepath.Join(t.TempDir(), "the-file")
old := parseTrusted(t, fmt.Sprintf(`{"declaration":1,"resources":[{"id":"m.old","type":"file","path":%q,`+
`"content":"port = 1\n"}]}`, path))
_, state, err := Apply(context.Background(), archHost(t), old, store.State{}, store.OriginDeclared, noServices,
nil, nil)
if err != nil {
t.Fatal(err)
}
was, _ := state.Find("m.old")
d := filledDecl(t, path, `"content":"","owner-at-runtime":"owned"`, "")
_, state, err = Apply(context.Background(), archHost(t), d, state, store.OriginDeclared, noServices, nil, nil)
if err != nil {
t.Fatal(err)
}
if got := readText(t, path); got != "port = 1\n" {
t.Errorf("the file handed to the client was rewritten or removed: %q", got)
}
if _, ok := state.Find("m.old"); ok {
t.Errorf("the record going was kept beside the filled file's")
}
if rec, ok := state.Find("m.conf"); !ok || rec.Wrote != was.Wrote || rec.Target != path {
t.Errorf("the filled file did not take the record of what the mesh wrote there: %+v", rec)
}
}
// elsewhere is a declaration with nothing at the filled path: one unrelated file.
func elsewhere(t *testing.T) *declaration.Declaration {
t.Helper()
return parseTrusted(t, fmt.Sprintf(`{"declaration":1,"resources":[{"id":"n.other","type":"file","path":%q,`+
`"content":"x\n"}]}`, filepath.Join(t.TempDir(), "other")))
}
internal/apply/filled_library_test.go
-57
View File
@@ -1,57 +0,0 @@
//go:build linux
package apply
import (
"context"
"fmt"
"path/filepath"
"strings"
"testing"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
)
// A container that reads a filled file and mounts a library whose disk is not there is held by the write verb as a
// full apply holds it (novox/hq ADR 0328 §9, issue 668): never created or recreated over the missing disk.
func TestAWriteDoesNotRecreateAContainerHeldForItsLibrary(t *testing.T) {
// No folder is given the group: the library's root is absent, and nothing is made.
gid := 61000
libraries(t, nil, nil)
base := t.TempDir()
root := filepath.Join(base, "gone", "media")
env := filepath.Join(base, "radarr.env")
raw := fmt.Sprintf(`{"declaration":1,"provisioner_account":"mesh-node-provisioner","filled_files":[%q],
"groups":[{"name":"mesh-media","gid":%d}],"libraries":[%s],"resources":[
{"id":"radarr.env","type":"file","path":%q,"content":"","mode":"0600","owner-at-runtime":"owned"},
{"id":"radarr.app","type":"container","name":"radarr","image":%q,"env-file":[%q],"volumes":[%q]}]}`,
env, gid, libJSON("media", root, gid, true, false, "movies"), env, pinned, env,
filepath.Join(root, "movies")+":/movies")
d, err := declaration.Parse([]byte(raw))
if err != nil {
t.Fatal(err)
}
var ran []string
run := func(_ context.Context, name string, args ...string) (string, error) {
ran = append(ran, name+" "+strings.Join(args, " "))
return "", nil
}
var state store.State
for _, content := range []string{"PORT=1\n", "PORT=2\n"} {
var report Report
report, state, err = WriteFilled(context.Background(), archHost(t), d, state, env, []byte(content), run,
nil, nil, nil)
if err != nil {
t.Fatal(err)
}
if o := outcomeOf(report, "radarr.app"); o.Action != "waiting" {
t.Errorf("after writing %q the container is %+v, not held for its library", content, o)
}
}
for _, line := range ran {
if strings.HasPrefix(line, "docker run") || strings.HasPrefix(line, "docker rm") {
t.Errorf("the container was created or removed over a missing disk: %v", ran)
}
}
}
internal/apply/filled_test.go
-271
View File
@@ -1,271 +0,0 @@
package apply
import (
"context"
"encoding/json"
"fmt"
"os"
"path/filepath"
"strings"
"testing"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
)
// Defends novox/hq ADR 0349 §4 (issue 668): the provisioner's client has the node-engine write a file it filled,
// and the node-engine writes it as an apply of the declared file with that content would — through applyFile,
// applyInto or applyBlock by who owns it at runtime — and makes what reads it take the change as an apply does.
// filledDecl is a declaration naming path as filled, with the file's fields after id, type and path, and more
// resources after it.
func filledDecl(t *testing.T, path, fields, more string) *declaration.Declaration {
t.Helper()
if more != "" {
more = "," + more
}
return parseTrusted(t, fmt.Sprintf(`{"declaration":1,"provisioner_account":"mesh-node-provisioner",
"filled_files":[%q],"resources":[{"id":"m.conf","type":"file","path":%q,%s}%s]}`, path, path, fields, more))
}
func TestAWriteOfAPathNotFilledIsRefused(t *testing.T) {
dir := t.TempDir()
filled := filepath.Join(dir, "filled.conf")
other := filepath.Join(dir, "other.conf")
d := parseTrusted(t, fmt.Sprintf(`{"declaration":1,"filled_files":[%q],"resources":[
{"id":"m.conf","type":"file","path":%q,"content":"x\n","owner-at-runtime":"owned"},
{"id":"m.other","type":"file","path":%q,"content":"y\n"}
]}`, filled, filled, other))
for _, path := range []string{other, filepath.Join(dir, "nothing.conf")} {
_, state, err := WriteFilled(context.Background(), archHost(t), d, store.State{}, path, []byte("evil\n"),
noServices, nil, nil, nil)
if err == nil || !strings.Contains(err.Error(), "does not name") {
t.Errorf("%s: a write of a path the declaration does not fill was not refused: %v", path, err)
}
if _, err := os.Stat(path); err == nil && path != other {
t.Errorf("%s was written", path)
}
if len(state.Resources) != 0 {
t.Errorf("a refused write recorded something: %+v", state.Resources)
}
}
if _, err := os.Stat(other); err == nil {
t.Errorf("%s was written", other)
}
}
// Each ownership is written by the write verb exactly as an apply of the declared file with the same content: the
// same action, the same bytes on disk, the same mode.
func TestAFilledFileIsWrittenAsItsOwnershipSays(t *testing.T) {
for _, c := range []struct {
name, fields, content, before, action string
}{
{"owned, absent", `"content":"","mode":"0640","owner-at-runtime":"owned"`, "port = 8080\n", "", "created"},
{"owned, over what is there", `"content":"","mode":"0640","owner-at-runtime":"owned"`, "port = 8080\n",
"port = 1\n", "updated"},
{"seed, absent", `"content":"","mode":"0600","owner-at-runtime":"seed"`, "user default on\n", "", "created"},
{"seed, present", `"content":"","mode":"0600","owner-at-runtime":"seed"`, "user default on\n",
"user default on\nuser app-one on\n", "kept"},
{"managed json", `"into":"json","content":"{}","owner-at-runtime":"managed"`,
`{"insecure-registries":["10.42.0.1:5000"]}`, machinesOwn, "updated"},
{"managed block", `"into":"block","content":"x\n","owner-at-runtime":"managed"`, "10.42.0.1 ace\n",
workstationHosts, "updated"},
} {
t.Run(c.name, func(t *testing.T) {
at := func() string {
path := filepath.Join(t.TempDir(), "the-file")
if c.before != "" {
if err := os.WriteFile(path, []byte(c.before), 0o644); err != nil {
t.Fatal(err)
}
}
return path
}
written := at()
report, state, err := WriteFilled(context.Background(), archHost(t), filledDecl(t, written, c.fields, ""),
store.State{}, written, []byte(c.content), noServices, nil, nil, nil)
if err != nil {
t.Fatal(err)
}
applied := at()
withContent := strings.Replace(c.fields, fmt.Sprintf(`"content":%q`, contentOf(t, c.fields)),
fmt.Sprintf(`"content":%q`, c.content), 1)
// The same file declared the old way: an apply writes a filled file only on the client's word (P7).
plain := parseTrusted(t, fmt.Sprintf(`{"declaration":1,"resources":[{"id":"m.conf","type":"file",`+
`"path":%q,%s}]}`, applied, withContent))
want, _, err := Apply(context.Background(), archHost(t), plain, store.State{}, store.OriginDeclared,
noServices, nil, nil)
if err != nil {
t.Fatal(err)
}
if got := report.Outcomes[0].Action; got != c.action || got != want.Outcomes[0].Action {
t.Errorf("written %s; an apply of the same content %s; want %s", got, want.Outcomes[0].Action, c.action)
}
if got, apply := readText(t, written), readText(t, applied); got != apply {
t.Errorf("written, the file holds\n%q\nan apply of the same content leaves\n%q", got, apply)
}
if a, b := modeAt(t, written), modeAt(t, applied); a != b {
t.Errorf("written mode %o, an apply's %o", a, b)
}
if rec, ok := state.Find("m.conf"); !ok || rec.Target != written || rec.Origin != store.OriginDeclared {
t.Errorf("the write is not recorded as the file's: %+v", rec)
}
})
}
}
func contentOf(t *testing.T, fields string) string {
t.Helper()
var holder struct {
Content string `json:"content"`
}
if err := json.Unmarshal([]byte("{"+fields+"}"), &holder); err != nil {
t.Fatal(err)
}
return holder.Content
}
func modeAt(t *testing.T, path string) os.FileMode {
t.Helper()
info, err := os.Stat(path)
if err != nil {
t.Fatal(err)
}
return info.Mode().Perm()
}
// A file that carried a secret keeps the mode a secret is written with: the client filled the value, and the file
// is no more readable for it.
func TestAFilledFileThatCarriedASecretIsNotWorldReadable(t *testing.T) {
path := filepath.Join(t.TempDir(), "token.env")
d := filledDecl(t, path, `"content":"TOKEN=${secret:token}\n","secrets":{"token":"c2VhbGVk"},"owner-at-runtime":"owned"`, "")
if _, _, err := WriteFilled(context.Background(), archHost(t), d, store.State{}, path, []byte("TOKEN=abc\n"),
noServices, nil, nil, nil); err != nil {
t.Fatal(err)
}
if m := modeAt(t, path); m != 0o600 {
t.Errorf("a filled file that carried a secret is mode %o", m)
}
}
func TestAWriteUnderReloadOnReloadsItsProgramOnce(t *testing.T) {
path := filepath.Join(t.TempDir(), "daemon.conf")
d := filledDecl(t, path, `"content":"","owner-at-runtime":"owned"`,
`{"id":"m.daemon","type":"service","unit":"m.service","state":"running","reload-on":["m.conf"]},
{"id":"m.other","type":"service","unit":"other.service","state":"running"}`)
var commands []string
var said []string
_, state, err := WriteFilled(context.Background(), archHost(t), d, store.State{}, path, []byte("level = 1\n"),
recordingServices(&commands), func(line string) { said = append(said, line) }, nil, nil)
if err != nil {
t.Fatal(err)
}
if n := count(commands, "systemctl reload m.service"); n != 1 {
t.Errorf("the program was reloaded %d times; commands were %v", n, commands)
}
joined := strings.Join(commands, "\n")
if strings.Contains(joined, "restart m.service") || strings.Contains(joined, "stop m.service") {
t.Errorf("the program was restarted, not reloaded: %v", commands)
}
if strings.Contains(joined, "other.service") {
t.Errorf("a service that does not read the file was touched: %v", commands)
}
if _, ok := state.Find("m.daemon"); !ok {
t.Errorf("the reloaded service was not recorded")
}
// The log says the path and the digest, never the content.
for _, line := range said {
if strings.Contains(line, "level = 1") {
t.Errorf("the content was said: %q", line)
}
}
if !strings.Contains(strings.Join(said, "\n"), digestOf("level = 1\n")) {
t.Errorf("the digest was not said: %q", said)
}
// The same content again moves nothing, and reloads nothing.
commands = nil
report, _, err := WriteFilled(context.Background(), archHost(t), d, state, path, []byte("level = 1\n"),
recordingServices(&commands), nil, nil, nil)
if err != nil {
t.Fatal(err)
}
if report.Outcomes[0].Action != "unchanged" || count(commands, "reload") != 0 {
t.Errorf("an unchanged write reloaded: %v %+v", commands, report.Outcomes)
}
}
func TestAWriteUnderRestartOnRestartsItsProgram(t *testing.T) {
path := filepath.Join(t.TempDir(), "daemon.conf")
d := filledDecl(t, path, `"content":"","owner-at-runtime":"owned"`,
`{"id":"m.daemon","type":"service","unit":"m.service","state":"running","restart-on":["m.conf"]}`)
var commands []string
if _, _, err := WriteFilled(context.Background(), archHost(t), d, store.State{}, path, []byte("level = 1\n"),
recordingServices(&commands), nil, nil, nil); err != nil {
t.Fatal(err)
}
// A restart is a stop and a start, once.
stops, starts := count(commands, "systemctl stop m.service"), count(commands, "systemctl start m.service")
if stops != 1 || starts != 1 {
t.Errorf("the program was stopped %d and started %d times; commands were %v", stops, starts, commands)
}
}
// A running container that reads the file at creation is recreated, as an apply recreates it for a changed file.
func TestAWriteOfAFileAContainerReadsRecreatesIt(t *testing.T) {
env := filepath.Join(t.TempDir(), "forge.env")
d := filledDecl(t, env, `"content":"","mode":"0600","owner-at-runtime":"owned"`,
`{"id":"forge.server","type":"container","name":"forge","image":"`+pinned+`","env-file":["`+env+`"]}`)
m := &machine{containers: map[string]*fakeContainer{}}
_, state, err := WriteFilled(context.Background(), archHost(t), d, store.State{}, env, []byte("PORT=5432\n"),
m.run, nil, nil, nil)
if err != nil {
t.Fatal(err)
}
if !m.did("docker run") {
t.Fatalf("the container was not created: %v", m.asked)
}
m.asked = nil
report, _, err := WriteFilled(context.Background(), archHost(t), d, state, env, []byte("PORT=5433\n"), m.run,
nil, nil, nil)
if err != nil {
t.Fatal(err)
}
if !m.removed("forge") || !m.did("docker run") {
t.Fatalf("the container kept what it read before the write: %v", m.asked)
}
if o := outcomeOf(report, "forge.server"); o.Action != "updated" || !strings.Contains(o.Detail, env) {
t.Errorf("the recreate did not say which file changed: %+v", o)
}
}
func count(lines []string, prefix string) int {
n := 0
for _, l := range lines {
if strings.Contains(l, prefix) {
n++
}
}
return n
}
// Two resources declaring one filled path — two modules' blocks in one file — are refused: the request carries the
// content only, so it cannot say whose block it is, and nothing is written.
func TestAFilledPathTwoResourcesDeclareIsRefused(t *testing.T) {
path := filepath.Join(t.TempDir(), "hosts")
if err := os.WriteFile(path, []byte(workstationHosts), 0o644); err != nil {
t.Fatal(err)
}
d := parseTrusted(t, fmt.Sprintf(`{"declaration":1,"filled_files":[%q],"resources":[
{"id":"a.hosts","type":"file","path":%q,"into":"block","content":"10.42.0.1 ace\n","owner-at-runtime":"managed"},
{"id":"b.hosts","type":"file","path":%q,"into":"block","content":"10.42.0.2 g14\n","owner-at-runtime":"managed"}
]}`, path, path, path))
_, state, err := WriteFilled(context.Background(), archHost(t), d, store.State{}, path, []byte("6.6.6.6 evil\n"),
noServices, nil, nil, nil)
if err == nil || !strings.Contains(err.Error(), "2 resources declare") {
t.Fatalf("a path two resources declare was not refused: %v", err)
}
if got := readText(t, path); got != workstationHosts || len(state.Resources) != 0 {
t.Errorf("something was written or recorded: %q %+v", got, state.Resources)
}
}
internal/apply/library.go
+6 -12
View File
@@ -318,17 +318,6 @@ func heldRoots(verdicts map[string]libraryVerdict, libs []declaration.Library) m
return out
}
// heldLibraries is the root of every library whose disk or mount is not there, with its name, judged now and read
// only: what a full apply holds its containers for (ADR 0328 §9), for one that judges without applying libraries.
func heldLibraries(ctx context.Context, libs []declaration.Library) map[string]string {
verdicts := map[string]libraryVerdict{}
for _, l := range libs {
rec, _ := libraryRecord(l.Library)
verdicts[l.Library] = assessLibrary(ctx, l, rec)
}
return heldRoots(verdicts, libs)
}
// mountsHeld is the library a container waits for: one whose root it mounts, or mounts below, is held.
func mountsHeld(c *declaration.Container, held map[string]string) string {
for _, v := range c.Volumes {
@@ -352,7 +341,12 @@ func WaitingOnLibrary(d *declaration.Declaration) map[string]bool {
if d == nil || len(d.Libraries) == 0 {
return out
}
held := heldLibraries(context.Background(), d.Libraries)
verdicts := map[string]libraryVerdict{}
for _, l := range d.Libraries {
rec, _ := libraryRecord(l.Library)
verdicts[l.Library] = assessLibrary(context.Background(), l, rec)
}
held := heldRoots(verdicts, d.Libraries)
for _, r := range d.Resources {
if c, ok := r.(*declaration.Container); ok && mountsHeld(c, held) != "" {
out[c.ID] = true
internal/apply/ownership_test.go
-129
View File
@@ -1,129 +0,0 @@
package apply
import (
"context"
"fmt"
"os"
"path/filepath"
"testing"
"github.com/novox/mesh-host/internal/store"
)
// Defends novox/hq ADR 0349 §5 (issue 667): a file that says who owns it at runtime is written as the
// form it names always was — owned as a whole file whose drift is written back, seed as create-once,
// managed as only its keys or its marked block. Each is applied beside its old form, through the same
// drift, and the two must end the same: the same actions, the same bytes on disk.
// ownershipCase is one ownership word and the old form it maps onto.
type ownershipCase struct {
name string
// said and old are the resource's fields after id, type and path: with the word, and in the old form.
said, old string
// before is what is at the path before the first apply; "" leaves the path absent.
before string
// drift is what the machine writes there between the two applies.
drift string
// first and second are the actions the old form is known to report.
first, second string
}
func TestEachOwnershipBehavesAsItsOldForm(t *testing.T) {
for _, c := range []ownershipCase{
{
name: "owned is a whole file whose drift is written back",
said: `"content":"correct\n","mode":"0644","owner-at-runtime":"owned"`,
old: `"content":"correct\n","mode":"0644"`,
drift: "someone edited this\n",
first: "created", second: "corrected",
},
{
name: "seed is written once and then the program's",
said: `"content":"user default on\n","mode":"0600","owner-at-runtime":"seed"`,
old: `"content":"user default on\n","mode":"0600","create-once":true`,
drift: "user default on\nuser app-one on\n",
first: "created", second: "kept",
},
{
name: "managed json sets only its keys",
said: `"into":"json","content":"{\"insecure-registries\":[\"10.42.0.1:5000\"]}","owner-at-runtime":"managed"`,
old: `"into":"json","content":"{\"insecure-registries\":[\"10.42.0.1:5000\"]}"`,
before: machinesOwn,
drift: `{"data-root":"/srv/other","log-opts":{"max-size":"10m"},"insecure-registries":["192.0.2.7:5000"]}`,
first: "updated", second: "corrected",
},
{
name: "managed block holds only its region",
said: `"into":"block","content":"10.42.0.1 ace\n","owner-at-runtime":"managed"`,
old: `"into":"block","content":"10.42.0.1 ace\n"`,
before: workstationHosts,
drift: workstationHosts + "\n" + marked("f", "6.6.6.6 evil\n") + "192.0.2.9 the-machines-own\n",
first: "updated", second: "corrected",
},
} {
t.Run(c.name, func(t *testing.T) {
saidActions, saidBytes := applyThroughDrift(t, c, c.said)
oldActions, oldBytes := applyThroughDrift(t, c, c.old)
if saidActions != oldActions {
t.Errorf("with the word the actions were %s; the old form's were %s", saidActions, oldActions)
}
if want := c.first + " " + c.second; oldActions != want {
t.Errorf("the old form's actions were %s; known to be %s", oldActions, want)
}
if saidBytes != oldBytes {
t.Errorf("with the word the file ends\n%q\nthe old form's ends\n%q", saidBytes, oldBytes)
}
})
}
}
// applyThroughDrift applies one file, lets the machine write the case's drift into it, applies again,
// and says the two actions and what the file holds at the end.
func applyThroughDrift(t *testing.T, c ownershipCase, fields string) (string, string) {
t.Helper()
path := filepath.Join(t.TempDir(), "the-file")
if c.before != "" {
if err := os.WriteFile(path, []byte(c.before), 0o644); err != nil {
t.Fatal(err)
}
}
d := parse(t, fmt.Sprintf(`{"declaration":1,"resources":[{"id":"f","type":"file","path":%q,%s}]}`, path, fields))
first, state, err := Apply(context.Background(), archHost(t), d, store.State{}, store.OriginDeclared, noServices, nil, nil)
if err != nil {
t.Fatal(err)
}
if err := os.WriteFile(path, []byte(c.drift), 0o644); err != nil {
t.Fatal(err)
}
second, _, err := Apply(context.Background(), archHost(t), d, state, store.OriginDeclared, noServices, nil, nil)
if err != nil {
t.Fatal(err)
}
return first.Outcomes[0].Action + " " + second.Outcomes[0].Action, readText(t, path)
}
// A seed said by the word alone, with no create-once beside it, is still kept once present: the word is
// enough, and the file the program grew is not written back.
func TestASeedByTheWordAloneKeepsWhatGrew(t *testing.T) {
path := filepath.Join(t.TempDir(), "acl.conf")
d := parse(t, fmt.Sprintf(`{"declaration":1,"resources":[
{"id":"acl","type":"file","path":%q,"content":"user default on\n","owner-at-runtime":"seed"}
]}`, path))
_, state, err := Apply(context.Background(), archHost(t), d, store.State{}, store.OriginDeclared, noServices, nil, nil)
if err != nil {
t.Fatal(err)
}
if err := os.WriteFile(path, []byte("grown\n"), 0o600); err != nil {
t.Fatal(err)
}
report, _, err := Apply(context.Background(), archHost(t), d, state, store.OriginDeclared, noServices, nil, nil)
if err != nil {
t.Fatal(err)
}
if got := report.Outcomes[0].Action; got != "kept" {
t.Errorf("a seed was %q", got)
}
if got := readText(t, path); got != "grown\n" {
t.Errorf("what grew in the seed was written over: %q", got)
}
}
internal/apply/vault_sealed_test.go
-112
View File
@@ -1,112 +0,0 @@
package apply
// A core credential the vault sealed to this machine (novox/hq ADR 0349 §16, issue 755): the controller carries the
// vault's box (ADR 0206) in the declaration and cannot open it; this machine opens it with its sealing key and writes
// the value, as it writes a value the controller sealed.
import (
"context"
"encoding/json"
"os"
"strings"
"testing"
"git.novox.be/novox/mesh-sdk/go/seal"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/identity"
"github.com/novox/mesh-host/internal/store"
)
// vaultSealedTo is what the vault answers the controller: the value in the ADR 0206 box, sealed to the machine's
// sealing key as SPKI PEM.
func vaultSealedTo(t *testing.T, key identity.SealingKey, value string) seal.Box {
t.Helper()
pub, err := key.PublicPEM()
if err != nil {
t.Fatal(err)
}
box, err := seal.Seal(value, pub)
if err != nil {
t.Fatal(err)
}
return box
}
func vaultSealedFile(t *testing.T, path string, box seal.Box) *declaration.Declaration {
t.Helper()
raw := map[string]any{"declaration": 1, "resources": []map[string]any{
{"id": "broker", "type": "file", "path": path, "vault-sealed": box},
}}
body, _ := json.Marshal(raw)
d, err := declaration.Parse(body)
if err != nil {
t.Fatal(err)
}
return d
}
func boxOpener(key identity.SealingKey) Unseal {
return func(sealed string) ([]byte, error) { return key.Open(sealed) }
}
func TestAFileTheVaultSealedToThisMachineIsOpenedAndWrittenUnreadableToOthers(t *testing.T) {
key, err := identity.GenerateSealingKey()
if err != nil {
t.Fatal(err)
}
path := t.TempDir() + "/broker"
d := vaultSealedFile(t, path, vaultSealedTo(t, key, "the-provisioners-bus-password"))
report, state, err := Apply(context.Background(), archHost(t), d, store.State{},
store.OriginCarried, noServices, nil, boxOpener(key))
if err != nil {
t.Fatal(err)
}
on, err := os.ReadFile(path)
if err != nil || string(on) != "the-provisioners-bus-password" {
t.Fatalf("the file holds %q, %v", on, err)
}
if info, _ := os.Stat(path); info.Mode().Perm() != 0o600 {
t.Errorf("a credential landed mode %o", info.Mode().Perm())
}
said, _ := json.Marshal(report)
kept, _ := json.Marshal(state)
for what, blob := range map[string][]byte{"the report": said, "the node's state": kept} {
if strings.Contains(string(blob), "the-provisioners-bus-password") {
t.Fatalf("%s carries the value in the clear", what)
}
}
}
func TestAVaultBoxSealedToAnotherMachineIsRefusedAndNothingWritten(t *testing.T) {
mine, _ := identity.GenerateSealingKey()
theirs, _ := identity.GenerateSealingKey()
path := t.TempDir() + "/broker"
d := vaultSealedFile(t, path, vaultSealedTo(t, theirs, "not for you"))
if _, _, err := Apply(context.Background(), archHost(t), d, store.State{},
store.OriginCarried, noServices, nil, boxOpener(mine)); err == nil {
t.Fatal("a box sealed to another machine was applied")
}
if _, err := os.Stat(path); err == nil {
t.Fatal("something was written before the failure")
}
}
func TestAVaultSealedFileSaysWhatIsInItOnce(t *testing.T) {
key, _ := identity.GenerateSealingKey()
box, _ := json.Marshal(vaultSealedTo(t, key, "x"))
for what, extra := range map[string]string{
"content": `"content":"a"`, "sealed": `"sealed":"b"`, "bytes": `"bytes":"Yw=="`,
"secrets": `"content":"${secret:x}","secrets":{"x":"c"}`, "into": `"into":"json","content":"{}"`,
} {
_, err := declaration.Parse([]byte(`{"declaration":1,"resources":[{"id":"f","type":"file","path":"/etc/x",` +
`"vault-sealed":` + string(box) + `,` + extra + `}]}`))
if err == nil {
t.Errorf("a vault-sealed file that also says %s was accepted", what)
}
}
if _, err := declaration.Parse([]byte(`{"declaration":1,"resources":[{"id":"f","type":"file","path":"/etc/x",` +
`"vault-sealed":{"v":2,"eph":"a","iv":"b","tag":"c","ct":"d"}}]}`)); err == nil {
t.Error("a box of a version the SDK does not open was accepted")
}
}
internal/bootstrap/bootstrap.go
+26 -60
View File
@@ -36,7 +36,6 @@ import (
"time"
"github.com/novox/mesh-host/internal/firewall"
"github.com/novox/mesh-host/internal/system"
"github.com/novox/mesh-host/internal/tunnel"
)
@@ -174,10 +173,6 @@ type Options struct {
// ToolsSource is where the shared base is built from. Everything with code of its own
// compiles against it, so it is the first thing the mesh builds for itself.
ToolsSource Source
// ToolsGoSource is where the Go toolchain is built from (`mesh-tools-go`, ADR 0142): every Go
// bundle compiles in it, and genesis's own modules have Go code (issue 616). Empty means the
// repository beside ToolsSource named mesh-tools-go, as the mesh keeps it.
ToolsGoSource Source
// CatalogSource is where the catalogue REPOSITORY is, for building its modules. The catalogue
// CHECKOUT (Catalogue, above) says what a module is; this is where a builder clones it.
CatalogSource Source
@@ -297,14 +292,12 @@ type Result struct {
Agent string `json:"host-agent,omitempty"`
// Registry is the mesh's own artifact store, once it answers.
Registry string `json:"registry,omitempty"`
RegistryReplied int `json:"registry-replied,omitempty"`
RegistryKnown bool `json:"registry-already-registered,omitempty"`
RegistryRunning string `json:"registry-container,omitempty"`
// RegistryTrusted are the names the runtime was told to trust the registry by, before the push.
RegistryTrusted []string `json:"registry-trusted,omitempty"`
PublishedAs string `json:"control-plane-image,omitempty"`
PublishedAlready bool `json:"control-plane-image-already-published,omitempty"`
Registry string `json:"registry,omitempty"`
RegistryReplied int `json:"registry-replied,omitempty"`
RegistryKnown bool `json:"registry-already-registered,omitempty"`
RegistryRunning string `json:"registry-container,omitempty"`
PublishedAs string `json:"control-plane-image,omitempty"`
PublishedAlready bool `json:"control-plane-image-already-published,omitempty"`
// Permanent is the control plane as an ordinary module.
Permanent string `json:"permanent-container,omitempty"`
@@ -562,9 +555,7 @@ func Run(ctx context.Context, o Options, d Deps, say func(string)) (Result, erro
// Named rather than attempted. Everything from step 6 on is a conversation with a
// control plane that a dry run has not raised, so there is nothing to ask and nothing
// honest to report about the answers.
say(" would then enrol " + o.Node + ", have the runtime trust the registry as " +
strings.Join(RegistryNames(o), ", ") + ",")
say(" install the registry from " +
say(" would then enrol " + o.Node + ", install the registry from " +
o.Catalogue + ", push the control plane's image into it,")
say(" reinstall the control plane as a module, and drop " +
rewritten.TempName)
@@ -663,10 +654,26 @@ func Run(ctx context.Context, o Options, d Deps, say func(string)) (Result, erro
return result, failed(StepOperator, err)
}
// ---- 7. registry, 8. publish -----------------------------------------------------------
published, err := RegistryAndPublish(ctx, o, d, sys, temporary, controlPlaneImage, &result, say)
// ---- 7. registry ----------------------------------------------------------------------
say("registry — somewhere for this mesh to keep its own images")
registry, err := InstallRegistry(ctx, o, d, temporary, say)
result.Registry, result.RegistryRunning = registry.Address, registry.Container
result.RegistryKnown, result.RegistryReplied = registry.Known, registry.Answered
if err != nil {
return result, err
return result, failed(StepRegistry, err)
}
// ---- 8. publish -----------------------------------------------------------------------
say("publish — the control plane's image gets its first manifest digest")
// **The image this mesh built, not the one the installer carried.** The carried one is the
// builder; publishing it here would put a builder in the registry under the control plane's
// name and install it as the control plane — which is exactly what happened the first time
// this ran, and presented as a control plane that started, printed a builder's usage, and
// exited cleanly over and over.
published, err := PublishControlPlane(ctx, o, d, controlPlaneImage, say)
result.PublishedAs, result.PublishedAlready = published.Reference, published.Already
if err != nil {
return result, failed(StepPublish, err)
}
// ---- 9. control plane -----------------------------------------------------------------
@@ -796,44 +803,3 @@ func Run(ctx context.Context, o Options, d Deps, say func(string)) (Result, erro
return result, nil
}
// RegistryAndPublish is steps 7 and 8: the runtime trusts the mesh's registry, the registry is
// installed and answers, and the control plane's image is pushed into it.
//
// **One function so the order is held by a test** (novox/hq issue 92). The runtime refuses plain
// HTTP to a registry it has not been told to trust, and the module that tells it is applied long
// after the push — so the trust is written and confirmed here, before anything is tagged or pushed.
// A refusal names the step it happened in.
func RegistryAndPublish(ctx context.Context, o Options, d Deps, sys system.System, temporary controlPlane,
controlPlaneImage string, result *Result, say func(string)) (Published, error) {
// ---- 7. registry ----------------------------------------------------------------------
say("registry — somewhere for this mesh to keep its own images")
// Trusted by the runtime before anything is pushed to it, by every name the mesh will use for
// it here: the module that writes this trust is applied long after the push (novox/hq issue 92).
trusted, err := TrustTheRegistry(ctx, o, sys, d.Run, say)
result.RegistryTrusted = trusted
if err != nil {
return Published{}, failed(StepRegistry, err)
}
registry, err := InstallRegistry(ctx, o, d, temporary, say)
result.Registry, result.RegistryRunning = registry.Address, registry.Container
result.RegistryKnown, result.RegistryReplied = registry.Known, registry.Answered
if err != nil {
return Published{}, failed(StepRegistry, err)
}
// ---- 8. publish -----------------------------------------------------------------------
say("publish — the control plane's image gets its first manifest digest")
// **The image this mesh built, not the one the installer carried.** The carried one is the
// builder; publishing it here would put a builder in the registry under the control plane's
// name and install it as the control plane — which is exactly what happened the first time
// this ran, and presented as a control plane that started, printed a builder's usage, and
// exited cleanly over and over.
published, err := PublishControlPlane(ctx, o, d, controlPlaneImage, say)
result.PublishedAs, result.PublishedAlready = published.Reference, published.Already
if err != nil {
return published, failed(StepPublish, err)
}
return published, nil
}
internal/bootstrap/builder_test.go
+21
View File
@@ -38,6 +38,27 @@ func recording(t *testing.T) (*controlRecorder, controlPlane) {
return c, controlPlane{container: "temp-mesh-controller", run: c.run, timeout: time.Second}
}
func TestTheBuildersPackageRegistryPortIsASettingAndNotTheManifest(t *testing.T) {
c, control := recording(t)
o := Options{Node: "anchor", Ports: FoundationPorts{Packages: 3100}, Wait: time.Second}
if _, err := installModule(context.Background(), o, control, BuilderModule,
[]byte(theBuildersBinding), quietly); err != nil {
t.Fatal(err)
}
if got := c.settings[BuilderModule+"-settings.json"]; got != `{"serves":{"port":3100}}` {
t.Errorf("the builder was told %q about the package registry's port", got)
}
// The manifest reaches the mesh as the catalogue wrote it. A port rewritten into it here is a
// port the next registration from the catalogue silently takes back.
if got := c.settings[BuilderModule+"-module.json"]; got != theBuildersBinding {
t.Errorf("the installer changed the builder's manifest:\n%s", got)
}
set, push := c.index("settings set "+BuilderModule), c.index("push anchor")
if set < 0 || push < 0 || set > push {
t.Fatalf("the port was not set before the push that raises the builder: %v", c.told)
}
}
func TestTheBuilderOnTheDefaultPackagesPortIsToldNothing(t *testing.T) {
c, control := recording(t)
o := Options{Node: "anchor", Wait: time.Second}
internal/bootstrap/bus_adoption.go
-179
View File
@@ -1,179 +0,0 @@
package bootstrap
import (
"context"
"crypto/sha256"
"encoding/hex"
"encoding/json"
"encoding/pem"
"errors"
"fmt"
"os"
"regexp"
"strconv"
"strings"
"time"
"github.com/novox/mesh-host/internal/declaration"
)
// Adopting the bus is the mesh's planned step for replacing it (novox/hq ADR 0333 §3, ADR 0236).
//
// The adoption replaces the bus's image once: the module's built image takes the place of the upstream
// one the bundle raised, on the same streams, users and certificate. The controller refuses a push that
// would replace the bus — "a planned step: `bus upgrade --why …` snapshots its streams first and checks
// them after" — so genesis stopped at step 18 (issue 590). The installer takes that step: it snapshots the
// streams itself, because at genesis no backup holder exists to be asked, and hands the snapshot to
// `bus upgrade`, which sends the machine, opens `bus-maintenance`, and ends the step done only once the
// bus answers healthy again — every stream, every durable consumer, a round trip (H-bus). The installer
// waits for that end.
// adoptionWhy is what the step is recorded as, for the record of hand acts and the bus's step.
const adoptionWhy = "genesis adopts its bus as the nats module (ADR 0333)"
// Where the installer keeps the streams' snapshot before the adoption, and the credential the snapshot
// program reads while it runs. Beside genesis's own credentials; variables so a test writes elsewhere.
var (
busSnapshotFile = "/var/lib/mesh-genesis/bus-before-adoption.tar"
busSnapshotCredential = "/var/lib/mesh-genesis/bus-snapshot.credential"
busCertificateFile = "/var/lib/mesh-broker-tls/tls.crt"
busControllerSecret = BusControllerFile
// busStepWithin bounds the wait for the controller to say the step ended: its own bound (15 min,
// mesh-controller busStepBound), and its doctor's next look after that.
busStepWithin = 17 * time.Minute
)
// adoptTheBus sends the nats module to this machine through the bus's planned step. A bus already on the
// module's image is the module's: it is sent like any module, and nothing is replaced.
func adoptTheBus(ctx context.Context, o Options, control controlPlane, bus *declaration.Container,
say func(string)) error {
if running := runningImage(ctx, control, bus.Name); running != "" && running != bus.Image {
say(" already adopted " + bus.Name + " — it runs the module's image")
_, err := pushNode(ctx, o, control, say)
return err
}
image, err := theModulesBusImage(ctx, o, control)
if err != nil {
return err
}
if err := snapshotTheBus(ctx, o, control, image, say); err != nil {
return err
}
say(" restarting " + bus.Name + " — once, onto the " + BusModule + " module's image, as the bus's " +
"planned step; its streams, users and certificate stay where they are (novox/hq ADR 0333 §3, ADR 0236)")
if _, err := control.within(2*o.Wait).tell(ctx, "bus", "upgrade", "--why", adoptionWhy,
"--reversible", "--snapshot-taken", busSnapshotFile); err != nil {
return fmt.Errorf("the bus's planned step was refused, so the bus is not replaced: %w", err)
}
return nil
}
// meshBuiltImage is an image the mesh built, as it records it: `artifact-store://<module>/<artifact>@sha256:…`.
var meshBuiltImage = regexp.MustCompile(`artifact-store://([a-z0-9][a-z0-9-]*/[a-z0-9][a-z0-9-]*@sha256:[0-9a-f]{64})`)
// theModulesBusImage is the nats module's server image, as the build just made it, pulled from this
// mesh's registry: the image that carries the snapshot program.
func theModulesBusImage(ctx context.Context, o Options, control controlPlane) (string, error) {
said, err := control.tell(ctx, "builds", BusModule, "-n", "1")
if err != nil {
return "", err
}
for _, m := range meshBuiltImage.FindAllStringSubmatch(said, -1) {
if strings.HasPrefix(m[1], BusModule+"/") {
return strings.TrimSuffix(o.Registry, "/") + "/" + m[1], nil
}
}
return "", fmt.Errorf("the mesh says no image it built for the %s module, so its snapshot program cannot "+
"be run before the bus is replaced:\n%s", BusModule, indent(strings.TrimSpace(said)))
}
// snapshotTheBus writes every stream of the running bus to busSnapshotFile with the module's own snapshot
// program (mesh-catalog modules/nats, ADR 0235), as the controller's account, pinned to the bus's
// certificate. The credential is written for the program's one run and removed after it.
func snapshotTheBus(ctx context.Context, o Options, control controlPlane, image string, say func(string)) error {
password, err := readCredentialFile(busControllerSecret)
if err != nil {
return fmt.Errorf("the controller's bus credential is not at %s, so the streams cannot be snapshotted "+
"before the bus is replaced: %w", busControllerSecret, err)
}
fingerprint, err := certificateFingerprint(busCertificateFile)
if err != nil {
return err
}
port := o.Ports.orDefaults().Bus
server := "tls://127.0.0.1:" + strconv.Itoa(port)
credential, err := json.Marshal(map[string]string{"url": server, "fingerprint": fingerprint,
"user": busControllerUser, "password": password})
if err != nil {
return err
}
if err := os.WriteFile(busSnapshotCredential, credential, 0o600); err != nil {
return fmt.Errorf("the snapshot's credential could not be written: %w", err)
}
defer os.Remove(busSnapshotCredential)
script := "docker run --rm -i --network host --entrypoint /usr/local/bin/mesh-nats-snapshot " + image +
" snapshot --server " + server + " --credential - < " + busSnapshotCredential + " > " + busSnapshotFile +
".partial && mv " + busSnapshotFile + ".partial " + busSnapshotFile
running, cancel := context.WithTimeout(ctx, busStepWithin)
defer cancel()
if out, err := control.run(running, "sh", "-c", script); err != nil {
_, _ = control.run(ctx, "rm", "-f", busSnapshotFile+".partial")
return fmt.Errorf("the bus's streams could not be snapshotted, so the bus is not replaced: %w\n%s",
err, indent(strings.TrimSpace(out)))
}
say(" snapshotted every stream of the bus, at " + busSnapshotFile + " (its module's snapshot program)")
return nil
}
// certificateFingerprint is the pin of the bus's certificate: sha256 of its DER, as every machine pins it.
func certificateFingerprint(path string) (string, error) {
raw, err := os.ReadFile(path)
if err != nil {
return "", fmt.Errorf("the bus's certificate is not at %s, so it cannot be pinned for the snapshot: %w", path, err)
}
block, _ := pem.Decode(raw)
if block == nil {
return "", fmt.Errorf("%s holds no PEM certificate", path)
}
sum := sha256.Sum256(block.Bytes)
return "sha256:" + hex.EncodeToString(sum[:]), nil
}
// waitForTheBusStep waits until the controller says the bus's planned step ended: done once the bus
// answered healthy — every stream, every durable consumer, a round trip — or failed, with its snapshot as
// the way back. The step's own bound is the controller's; this waits a little past it.
func waitForTheBusStep(ctx context.Context, o Options, control controlPlane, say func(string)) error {
deadline := time.Now().Add(busStepWithin)
var last string
for {
said, err := control.tell(ctx, "bus")
if err == nil {
for _, line := range strings.Split(said, "\n") {
if strings.HasPrefix(strings.TrimSpace(line), "last step ") {
last = strings.TrimSpace(line)
}
}
switch {
case strings.Contains(last, ": done at "):
say(" bus step done " + last)
return nil
case strings.Contains(last, ": failed at "):
return fmt.Errorf("the bus's planned step ended failed: %s\n%s", last, indent(strings.TrimSpace(said)))
}
}
if !time.Now().Before(deadline) {
break
}
select {
case <-ctx.Done():
return ctx.Err()
case <-time.After(answerEvery):
}
}
if last == "" {
return errors.New("the controller says no bus step after the adoption was sent; `mesh-controller bus` says how it stands")
}
return fmt.Errorf("the bus's planned step did not end within %s: %s", busStepWithin, last)
}
internal/bootstrap/bus_adoption_test.go
-135
View File
@@ -1,135 +0,0 @@
package bootstrap
import (
"context"
"encoding/json"
"os"
"strings"
"testing"
"time"
)
// Defends novox/hq issue 590 at its broker step (ADR 0333 §3, ADR 0236): the adoption replaces the bus's
// image, and the mesh replaces its bus only as a planned step. A plain push was refused — "sending anchor
// would replace the bus … a planned step: `bus upgrade --why …`" — and genesis stopped at step 18.
func TestTheBrokerStepAdoptsTheBusThroughItsPlannedStep(t *testing.T) {
catalogue := aCatalogue(t, map[string]string{BusModule: aBusModule(theModulesLayout)})
rewritten, err := Rewrite(theRealBundle(t), held)
if err != nil {
t.Fatal(err)
}
bundleImage := containerNamed(rewritten.Declaration, "mesh-broker-nats").Image
upgraded := false
var credential []byte
runtime := &asked{}
agrees := aMeshThatAgrees(nil)
runtime.answer = func(name string, args []string) (string, error) {
joined := strings.Join(args, " ")
switch {
case name == "sh" && strings.Contains(joined, "mesh-nats-snapshot"):
credential, _ = os.ReadFile(busSnapshotCredential) // as the program reads it
case strings.Contains(joined, "bus upgrade --why"):
upgraded = true
case strings.Contains(joined, "inspect --format {{.Config.Image}} mesh-broker-nats"):
if upgraded {
return theModulesImage + "\n", nil
}
return bundleImage + "\n", nil
}
return agrees(name, args)
}
_, fetch := aBusThatAdopts()
o := installing(t, catalogue)
o.CatalogSource = Source{Repository: "git://forge/mesh-catalog.git"}
o.Wait = 5 * time.Second
said, err := busStepWith(t, o, runtime, fetch)
if err != nil {
t.Fatal(err)
}
snapshot, upgrade := -1, -1
for i, c := range runtime.commands {
switch {
case strings.HasPrefix(c, "sh -c") && strings.Contains(c, "mesh-nats-snapshot"):
snapshot = i
for _, want := range []string{"127.0.0.1:5000/nats/server@sha256:" + strings.Repeat("4", 64),
"snapshot --server tls://127.0.0.1:5671 --credential - < " + busSnapshotCredential,
"> " + busSnapshotFile + ".partial && mv " + busSnapshotFile + ".partial " + busSnapshotFile} {
if !strings.Contains(c, want) {
t.Errorf("the snapshot is not taken with %q: %s", want, c)
}
}
case strings.Contains(c, "/mesh-controller bus upgrade"):
upgrade = i
if want := "bus upgrade --why " + adoptionWhy + " --reversible --snapshot-taken " + busSnapshotFile; !strings.HasSuffix(c, want) {
t.Errorf("the planned step is asked as %q; want it to end %q", c, want)
}
case strings.HasSuffix(c, "/mesh-controller push anchor"):
t.Errorf("the bus's module was sent by a plain push, which the mesh refuses for the bus: %v", runtime.commands)
}
}
if snapshot < 0 || upgrade < 0 || snapshot > upgrade {
t.Fatalf("the streams were not snapshotted before the bus's planned step (snapshot %d, upgrade %d): %v",
snapshot, upgrade, runtime.commands)
}
var cred map[string]string
if err := json.Unmarshal(credential, &cred); err != nil {
t.Fatalf("the snapshot program was given no credential: %v", err)
}
fingerprint, _ := certificateFingerprint(busCertificateFile)
if cred["user"] != "controller" || cred["password"] != "BUS-PW" || cred["fingerprint"] != fingerprint ||
cred["url"] != "tls://127.0.0.1:5671" {
t.Errorf("the snapshot's credential is %v", cred)
}
if _, err := os.Stat(busSnapshotCredential); !os.IsNotExist(err) {
t.Error("the snapshot's credential was left on the machine")
}
text := strings.Join(said, "\n")
for _, want := range []string{"snapshotted every stream", "restarting mesh-broker-nats — once",
"bus answers mesh-broker-nats", "bus step done last step 1", "adopted mesh-broker-nats"} {
if !strings.Contains(text, want) {
t.Errorf("the step did not say %q:\n%s", want, text)
}
}
}
// A planned step the controller ends failed is the broker step failing, with the controller's words.
func TestABusStepThatEndsFailedIsRefused(t *testing.T) {
aGenesisBusOnDisk(t)
runtime := &asked{answer: func(name string, args []string) (string, error) {
return "last step 1: not known → 53e51864 on anchor by installer (x): failed at 2026-10-13 11:59; snapshot s\n" +
" stream KV_x has no messages\n", nil
}}
control := controlPlane{container: "mesh-controller", run: runtime.run, timeout: time.Second}
err := waitForTheBusStep(context.Background(), Options{}, control, quietly)
if err == nil || !strings.Contains(err.Error(), "failed at") || !strings.Contains(err.Error(), "KV_x") {
t.Fatalf("a failed bus step was not refused with its words: %v", err)
}
}
// The bundle's bus runs the release the module's image is built on, pinned by the same digest, so the
// adoption changes no server release and is reversible as the step says (ADR 0333 §3).
func TestTheBundlesBusRunsTheReleaseTheModulesImageIsBuiltOn(t *testing.T) {
var m struct {
Build struct {
On []struct{ Arg, Image string } `json:"on"`
} `json:"build"`
}
if err := json.Unmarshal(theCataloguesModule(t, BusModule), &m); err != nil {
t.Fatal(err)
}
base := ""
for _, on := range m.Build.On {
if on.Arg == "NATS_BASE" {
base = on.Image
}
}
rewritten, err := Rewrite(theRealBundle(t), held)
if err != nil {
t.Fatal(err)
}
if got := containerNamed(rewritten.Declaration, "mesh-broker-nats").Image; got != base {
t.Errorf("the bundle raises the bus on %s; the nats module's image is built on %s", got, base)
}
}
internal/bootstrap/genesis_registry_test.go
-65
View File
@@ -1,65 +0,0 @@
package bootstrap
import (
"context"
"encoding/json"
"strings"
"testing"
"time"
)
// Genesis seals the build agent the forge it raised (novox/hq ADR 0336 §3, issue 121): the agent's
// npm-package-registry requirement is optional and binds only once the gitea module is assigned, after
// the base, so until then the packages step's genesis document is how the sdk and base steps reach the
// forge.
func TestThePackagesStepSealsTheGenesisDocumentToTheBuildAgent(t *testing.T) {
if BuilderModule != "build-agent" {
t.Fatalf("genesis installs %q; the document is the build agent's", BuilderModule)
}
c, control := recording(t)
o := Options{Node: "anchor", Ports: FoundationPorts{Packages: 3100}, Wait: time.Second}
doc := genesisRegistry(o.Ports.orDefaults(), "builder-password")
if err := deliverGenesisRegistry(context.Background(), o, control, doc, quietly); err != nil {
t.Fatal(err)
}
carried := c.settings["accepting-"+GenesisRegistrySecret]
var got GenesisRegistry
if err := json.Unmarshal([]byte(carried), &got); err != nil {
t.Fatalf("the genesis document carried is not JSON: %q", carried)
}
want := GenesisRegistry{Scheme: "http", At: "127.0.0.1", Port: 3100, NpmPath: "/api/packages/novox/npm/",
As: "mesh-builder", Password: "builder-password"}
if got != want {
t.Fatalf("the genesis document says %+v, want %+v", got, want)
}
accept := c.index("secret accept anchor build-agent genesis-package-registry --from /accepting-genesis-package-registry")
push := c.index("push anchor")
if accept < 0 || push < 0 || accept > push {
t.Fatalf("the genesis document was not sealed to the build agent before the push: %v", c.told)
}
}
// What the retired builder was given and the build agent declares neither of: an `npm-password`
// own-secret and a `serves` setting for its carried binding (ADR 0336 §3).
func TestNoStepGivesTheBuildAgentAServesSettingOrAnNpmPassword(t *testing.T) {
for _, module := range []string{BuilderModule} {
c, control := recording(t)
o := Options{Node: "anchor", Ports: FoundationPorts{Packages: 3100}, Wait: time.Second}
if _, err := installModule(context.Background(), o, control, module, []byte(`{"module":"`+module+`"}`), quietly); err != nil {
t.Fatal(err)
}
if got, set := c.settings[module+"-settings.json"]; set {
t.Errorf("%s was given a setting on a node with a non-default packages port: %s", module, got)
}
if err := deliverGenesisRegistry(context.Background(), o, control,
genesisRegistry(o.Ports.orDefaults(), "p"), quietly); err != nil {
t.Fatal(err)
}
for _, told := range c.told {
if strings.Contains(told, "npm-password") || strings.Contains(told, "serves") {
t.Errorf("genesis told the controller %q", told)
}
}
}
}
internal/bootstrap/genesis_toolchains_test.go
-89
View File
@@ -1,89 +0,0 @@
package bootstrap
import (
"context"
"encoding/json"
"strings"
"testing"
"time"
)
// Defends novox/hq issue 616: genesis's base step builds the toolchain of every language a module
// genesis installs is compiled in. It built only mesh-tools, and the store's module, with Go code,
// stopped genesis at step 17: "this mesh holds no copy of it. Build mesh-tools-go first".
// genesisModules is every catalogue module genesis installs on its own, read from the steps that
// install them: the registry, the build agent, the store, the bus, the vault, the catalogue and the
// filter (extras are the operator's choice and come after the base too, but are not genesis's own).
var genesisModules = []string{RegistryModule, BuilderModule, "postgres", BusModule, "mesh-vault", "mesh-catalog", "nftables"}
// Every genesis module's bundles are compiled in a toolchain the base step builds, read from the real
// catalogue's manifests: a module that gains a language genesis does not build fails here, not at a
// genesis.
func TestTheBaseStepBuildsEveryToolchainAGenesisModuleIsCompiledIn(t *testing.T) {
built := map[string]bool{}
for _, tc := range genesisToolchains {
built[tc.Language] = true
}
for _, module := range genesisModules {
var m struct {
Build struct {
Artifacts []struct {
Name, Kind, Language string
} `json:"artifacts"`
} `json:"build"`
}
if err := json.Unmarshal(theCataloguesModule(t, module), &m); err != nil {
t.Fatalf("%s: %v", module, err)
}
for _, a := range m.Build.Artifacts {
if a.Kind != "bundle" {
continue // an image builds on the bases its Dockerfile names, pinned upstream
}
if !built[a.Language] {
t.Errorf("%s's %q is a %s bundle, and genesis's base step builds no %s toolchain",
module, a.Name, a.Language, a.Language)
}
}
}
}
// And the base step asks for each toolchain's build: the base, then the Go toolchain beside it.
func TestTheBaseStepBuildsTheGoToolchainBesideTheBase(t *testing.T) {
runtime := &asked{answer: aMeshThatAgrees(nil)}
control := controlPlane{container: "mesh-controller", run: runtime.run, timeout: time.Second}
o := Options{ToolsSource: Source{Repository: "git://192.0.2.30/mesh-tools.git"}}
if err := BuildBase(context.Background(), o, control, quietly); err != nil {
t.Fatal(err)
}
var builds []string
for _, c := range runtime.commands {
if strings.Contains(c, "/mesh-controller build ") {
builds = append(builds, c[strings.Index(c, "build "):])
}
}
want := []string{
"build git://192.0.2.30/mesh-tools.git --ref main --wait 1200s",
"build git://192.0.2.30/mesh-tools-go.git --ref main --wait 1200s",
}
if strings.Join(builds, "\n") != strings.Join(want, "\n") {
t.Fatalf("the base step built\n%s\nand not\n%s", strings.Join(builds, "\n"), strings.Join(want, "\n"))
}
}
func TestTheGoToolchainIsAsGivenOrBesideTheBase(t *testing.T) {
for _, c := range []struct{ tools, given, want string }{
{"http://forge/novox/mesh-tools.git", "", "http://forge/novox/mesh-tools-go.git"},
{"http://forge/novox/mesh-tools", "", "http://forge/novox/mesh-tools-go"},
{"http://forge/other/base.git", "http://forge/other/go.git", "http://forge/other/go.git"},
} {
got, err := toolsGoSource(Options{ToolsSource: Source{Repository: c.tools}, ToolsGoSource: Source{Repository: c.given}})
if err != nil || got.Repository != c.want {
t.Errorf("tools %s, given %q: %q %v; want %s", c.tools, c.given, got.Repository, err, c.want)
}
}
if _, err := toolsGoSource(Options{ToolsSource: Source{Repository: "http://forge/other/base.git"}}); err == nil ||
!strings.Contains(err.Error(), "--tools-go-source") {
t.Errorf("a base not named mesh-tools, and no Go toolchain given, was not refused naming the flag: %v", err)
}
}
internal/bootstrap/module.go
+2 -34
View File
@@ -137,26 +137,7 @@ func registerAndAssign(ctx context.Context, o Options, control controlPlane, mod
// should hold, seals every secret in them and publishes them, and on a first node that is the
// slowest thing the mesh does.
func pushNode(ctx context.Context, o Options, control controlPlane, say func(string)) (string, error) {
return pushing(ctx, o, control, false, say)
}
// pushNodeApplied is pushNode that returns only once the node reports it applied exactly what it was
// sent (`push <node> --wait`, novox/hq ADR 0010): for a push whose content the very next step reads
// on the machine. A plain push returns as soon as the declaration is sent, and the node applies it a
// moment later (issue 615).
func pushNodeApplied(ctx context.Context, o Options, control controlPlane, say func(string)) (string, error) {
return pushing(ctx, o, control, true, say)
}
func pushing(ctx context.Context, o Options, control controlPlane, applied bool, say func(string)) (string, error) {
args := []string{"push", o.Node}
patience := o.Wait
if applied {
args = append(args, "--wait", o.Wait.String())
// The controller's own wait, and time to compose and send before it starts.
patience = 2 * o.Wait
}
said, err := control.within(patience).tell(ctx, args...)
said, err := control.within(o.Wait).tell(ctx, "push", o.Node)
if err != nil {
return "", fmt.Errorf(
"%w\n\nWhat was registered and assigned is registered and assigned, and this node has "+
@@ -164,20 +145,7 @@ func pushing(ctx context.Context, o Options, control controlPlane, applied bool,
"at all. Fix what it named and run this installer again: it will find the module "+
"already registered and try the push again", err)
}
if applied {
say(" pushed " + o.Node + ", and it applied it")
} else {
say(" pushed " + o.Node)
}
// **Every push may carry a new bus user, and until the adoption the bus hears of it only from here**
// (novox/hq issue 608, ADR 0331 §5, ADR 0333 §4). A module's account is minted into the mesh's
// records and sealed to this node with the push; the bus reads its user list from a file, and until
// the broker step adopts the bus as the nats module nothing but this installer writes that file. The
// list was placed only when the node enrolled, so the build agent's account (step 13) never reached
// the bus and the agent was refused ("Authorization Violation") until the sdk step gave up.
if err := placeUntilAdopted(ctx, control, say); err != nil {
return "", err
}
say(" pushed " + o.Node)
return strings.TrimSpace(said), nil
}
internal/bootstrap/phase2.go
+5 -47
View File
@@ -25,62 +25,20 @@ import (
// buildWait bounds one module build. Generous, because the first build compiles a toolchain.
const buildWait = 20 * time.Minute
// BuildBase asks the mesh to build the shared base every module with code of its own stands on, and
// the toolchain of every language genesis's own modules are written in.
// BuildBase asks the mesh to build the shared base every module with code of its own stands on.
//
// **First, because until it exists nothing else with code can be built.** Not registered as a
// module here: it is never assigned — it runs nowhere — and the build itself records what was
// made, which is all anything downstream reads.
//
// **Every toolchain a genesis module compiles in** (novox/hq issue 616). A bundle is compiled in its
// language's base (mesh-controller internal/builder toolchain.go): TypeScript in mesh-tools, Go in
// mesh-tools-go. The store's and the bus's modules carry Go code, so a genesis that built only
// mesh-tools stopped at the store step: "this mesh holds no copy of it. Build mesh-tools-go first".
// The bases are the ones genesisToolchains names, and a test holds them to the languages of every
// module genesis installs, read from the catalogue.
func BuildBase(ctx context.Context, o Options, control controlPlane, say func(string)) error {
if o.ToolsSource.Repository == "" {
return fmt.Errorf("phase two needs --tools-source: the shared base is built from its " +
"own repository, and an installer told nothing cannot know where that is")
}
goSource, err := toolsGoSource(o)
if err != nil {
return err
}
for _, src := range []Source{o.ToolsSource, goSource} {
say(" building " + src.Repository + " at " + refOr(src.Ref))
if _, err := control.within(buildWait).tell(ctx,
"build", src.Repository, "--ref", refOr(src.Ref), "--wait", "1200s"); err != nil {
return err
}
}
return nil
}
// genesisToolchains is the base each language's bundles compile in, for the languages genesis's own
// modules are written in — the same pairs as mesh-controller internal/builder toolchain.go — in the
// order BuildBase builds them.
var genesisToolchains = []struct{ Language, Base string }{
{"typescript", "mesh-tools"},
{"go", "mesh-tools-go"},
}
// toolsGoSource is where the Go toolchain is built from: as given, or the repository beside the base's
// named mesh-tools-go, which is how the mesh keeps its repositories side by side.
func toolsGoSource(o Options) (Source, error) {
if o.ToolsGoSource.Repository != "" {
return o.ToolsGoSource, nil
}
repo := strings.TrimSuffix(o.ToolsSource.Repository, "/")
suffix := ""
if strings.HasSuffix(repo, ".git") {
repo, suffix = strings.TrimSuffix(repo, ".git"), ".git"
}
if !strings.HasSuffix(repo, "mesh-tools") {
return Source{}, fmt.Errorf("the Go toolchain is built from mesh-tools-go, and --tools-source (%s) is not a "+
"repository named mesh-tools to find it beside; give --tools-go-source", o.ToolsSource.Repository)
}
return Source{Repository: repo + "-go" + suffix, Ref: o.ToolsGoSource.Ref}, nil
say(" building " + o.ToolsSource.Repository + " at " + refOr(o.ToolsSource.Ref))
_, err := control.within(buildWait).tell(ctx,
"build", o.ToolsSource.Repository, "--ref", refOr(o.ToolsSource.Ref), "--wait", "1200s")
return err
}
// InstallFromCatalogue builds a catalogue module and installs it on this machine.
internal/bootstrap/phase3.go
+22 -16
View File
@@ -84,7 +84,7 @@ func InstallStore(ctx context.Context, o Options, control controlPlane,
say); err != nil {
return err
}
say(" adopted " + store.Name + " — the foundation's store is now the " + module + " module")
say(" adopted mesh-store — the foundation's store is now the " + module + " module")
return nil
}
@@ -93,8 +93,7 @@ func InstallStore(ctx context.Context, o Options, control controlPlane,
// be in it. InstallFromCatalogue is the same sequence without that moment; the store, the broker
// and the vault each need it or need the shape, and three copies of it drifted.
func installProvider(ctx context.Context, o Options, control controlPlane, module string,
manifest []byte, buildNote string, beforePush func() error, say func(string),
send ...func() error) error {
manifest []byte, buildNote string, beforePush func() error, say func(string)) error {
remote := "/" + module + "-module.json"
if err := control.carrying(ctx, module+"-module.json", manifest, remote); err != nil {
@@ -133,11 +132,6 @@ func installProvider(ctx context.Context, o Options, control controlPlane, modul
return err
}
}
// A module whose first send is not a plain push says how it is sent: the bus, whose replacement is a
// planned step (ADR 0236), is sent by that step (adoptTheBus).
if len(send) > 0 && send[0] != nil {
return send[0]()
}
_, err := pushNode(ctx, o, control, say)
return err
}
@@ -182,10 +176,9 @@ const BusModule = "nats"
// **The image is the one thing that changes, once** (ADR 0333 §3). The store's module pins the upstream
// postgres the foundation runs; the bus's module builds its server image (it carries the mesh's snapshot
// tool), so no pinned reference can be equal to the foundation's. The applier recreates the container on
// the module's image, on the same streams, users and certificate. The mesh replaces its bus only as a
// planned step (ADR 0236), so the module is sent by that step: its streams snapshotted first, the restart
// said, and the installer waits until the container runs the module's image, the bus answers, and the
// controller ends the step done (bus_adoption.go), so nothing after this step dials a bus that is restarting.
// the module's image, on the same streams, users and certificate. This says so before the push and waits
// after it until the container runs the module's image and the bus answers, so nothing after this step
// dials a bus that is restarting.
func InstallBroker(ctx context.Context, o Options, control controlPlane,
foundation *declaration.Declaration, fetch func(context.Context, string) (int, string, error),
say func(string)) error {
@@ -204,15 +197,20 @@ func InstallBroker(ctx context.Context, o Options, control controlPlane,
}
say(" adopting " + bus.Name + " — the bus the foundation raised, its streams kept")
if err := installProvider(ctx, o, control, module, manifest, "(the server, adopting the foundation's bus)",
nil, say, func() error { return adoptTheBus(ctx, o, control, bus, say) }); err != nil {
func() error {
if running := runningImage(ctx, control, bus.Name); running != "" && running != bus.Image {
say(" already restarted " + bus.Name + " — it runs the module's image")
return nil
}
say(" restarting " + bus.Name + " — once, onto the " + module + " module's image; its " +
"streams, users and certificate stay where they are (novox/hq ADR 0333 §3)")
return nil
}, say); err != nil {
return err
}
if err := waitForTheAdoptedBus(ctx, o, control, bus, fetch, say); err != nil {
return err
}
if err := waitForTheBusStep(ctx, o, control, say); err != nil {
return err
}
say(" adopted " + bus.Name + " — the foundation's bus is now the " + module +
" module, and its user list is the module's from here on")
return nil
@@ -480,3 +478,11 @@ func serverMatchesFoundation(manifest []byte, store *declaration.Container, modu
"store with. Its server container has to carry the name the foundation raised",
module, store.Name)
}
// orUnknown is s, or a word for nothing: trust.go held it, and this branch takes trust.go out.
func orUnknown(s string) string {
if strings.TrimSpace(s) == "" {
return "(nothing)"
}
return s
}
internal/bootstrap/phase3_test.go
+2 -30
View File
@@ -69,7 +69,6 @@ func aBusThatAdopts() (*asked, func(context.Context, string) (int, string, error
func busStepWith(t *testing.T, o Options, runtime *asked,
fetch func(context.Context, string) (int, string, error)) ([]string, error) {
t.Helper()
aGenesisBusOnDisk(t)
rewritten, err := Rewrite(theRealBundle(t), held)
if err != nil {
t.Fatal(err)
@@ -223,7 +222,7 @@ func TestTheBrokerStepSaysTheRestartAndWaitsForTheBusToAnswer(t *testing.T) {
runtime.answer = func(name string, args []string) (string, error) {
joined := strings.Join(args, " ")
switch {
case strings.Contains(joined, "bus upgrade --why"):
case strings.Contains(joined, "push anchor"):
pushed = true
case strings.Contains(joined, "inspect --format {{.Config.Image}} mesh-broker-nats"):
if !pushed {
@@ -305,7 +304,7 @@ func TestARerunAfterTheAdoptionSaysTheBusIsTheModulesAlready(t *testing.T) {
t.Fatal(err)
}
text := strings.Join(said, "\n")
if strings.Contains(text, "restarting") || !strings.Contains(text, "already adopted mesh-broker-nats") {
if strings.Contains(text, "restarting") || !strings.Contains(text, "already restarted mesh-broker-nats") {
t.Errorf("a re-run announced a restart, or did not say the bus is the module's already:\n%s", text)
}
}
@@ -418,30 +417,3 @@ func TestTheBundlesBusFilesAreSeeds(t *testing.T) {
}
}
}
// aGenesisBusOnDisk puts what the adoption reads on the machine in a test's own directory: the bus's
// certificate, the controller's bus credential, and where the snapshot goes. The step's wait is short.
func aGenesisBusOnDisk(t *testing.T) {
t.Helper()
dir := t.TempDir()
cert := filepath.Join(dir, "tls.crt")
if err := os.WriteFile(cert, []byte(aCertificatePEM), 0o644); err != nil {
t.Fatal(err)
}
secret := filepath.Join(dir, "bus-controller.secret")
if err := os.WriteFile(secret, []byte("BUS-PW"), 0o600); err != nil {
t.Fatal(err)
}
was := []string{busSnapshotFile, busSnapshotCredential, busCertificateFile, busControllerSecret}
wasWithin := busStepWithin
busSnapshotFile, busSnapshotCredential = filepath.Join(dir, "bus.tar"), filepath.Join(dir, "bus.credential")
busCertificateFile, busControllerSecret = cert, secret
busStepWithin = 50 * time.Millisecond
t.Cleanup(func() {
busSnapshotFile, busSnapshotCredential, busCertificateFile, busControllerSecret = was[0], was[1], was[2], was[3]
busStepWithin = wasWithin
})
}
// aCertificatePEM is a PEM block standing for the bus's certificate; its pin is the sha256 of these bytes.
const aCertificatePEM = "-----BEGIN CERTIFICATE-----\nAAECAwQFBgcICQ==\n-----END CERTIFICATE-----\n"
internal/bootstrap/phase_packages.go
+17 -60
View File
@@ -2,7 +2,6 @@ package bootstrap
import (
"context"
"encoding/json"
"fmt"
"net/http"
"os"
@@ -25,7 +24,7 @@ import (
const (
// foundationStore is the foundation's postgres container — the mesh's own memory, raised from the
// bundle. gitea's bootstrap database lives here too, so a mesh runs one postgres (issue 051).
foundationStore = "postgres"
foundationStore = "mesh-store"
// giteaBootstrap is the gitea server raised directly at genesis, before gitea is a module —
// under the name the gitea MODULE declares for its container, so the module finds it and holds
// it rather than raising a second forge beside it (novox/hq ADR 0163, rule 7; issue 090).
@@ -50,8 +49,8 @@ const (
packagesTeam = "packages"
// giteaAdminUser is the admin the bootstrap creates and the provisioner later authenticates as.
giteaAdminUser = "mesh-admin"
// builderGiteaUser is the gitea account the build agent publishes and pulls with at genesis: the
// `as` of the genesis document (novox/hq ADR 0336 §3). Retiring it is issue 602.
// builderGiteaUser is the gitea account the builder publishes and pulls with at genesis. It is
// the `as` the builder's static package binding names.
builderGiteaUser = "mesh-builder"
// giteaDBRole/giteaDBName is gitea's own database in the foundation store.
giteaDBRole = "mesh_gitea"
@@ -134,8 +133,8 @@ func RaisePackageRegistry(ctx context.Context, o Options, d Deps, control contro
return err
}
say(" sealing the build agent the genesis document for the forge it raised")
if err := deliverGenesisRegistry(ctx, o, control, genesisRegistry(ports, builderPassword), say); err != nil {
say(" delivering the builder its registry credential")
if err := deliverBuilderNpm(ctx, o, control, builderPassword, say); err != nil {
return err
}
@@ -363,69 +362,27 @@ func packagePasswords() (db, admin, builder string, err error) {
return db, admin, builder, nil
}
// GenesisRegistrySecret is the build agent's own-secret the genesis document is sealed as (novox/hq
// ADR 0336 §3). The build-agent manifest declares it, issued outside the mesh: only this installer
// gives one, on genesis's node.
const GenesisRegistrySecret = "genesis-package-registry"
// GenesisRegistry is the genesis document (novox/hq ADR 0336 §3): the forge this step raised, the port
// this node gave the package registry, its npm path, and the account this step made with its password.
// mesh-controller's cmd/mesh-builder reads it under the same field names while the build agent's
// npm-package-registry requirement is unbound, and prefers the binding once the gitea module is
// assigned and it binds.
type GenesisRegistry struct {
Scheme string `json:"scheme"`
At string `json:"at"`
Port int `json:"port"`
NpmPath string `json:"npm-path"`
As string `json:"as"`
Password string `json:"password"`
}
// genesisRegistry is the document for the forge raised here: on the machine's own loopback, where it
// answers on the host network and where its ROOT_URL says it is (raiseGiteaServer), on the port this
// node gave the package registry.
func genesisRegistry(ports FoundationPorts, password string) GenesisRegistry {
return GenesisRegistry{Scheme: "http", At: "127.0.0.1", Port: ports.Packages,
NpmPath: "/api/packages/" + packagesOrg + "/npm/", As: builderGiteaUser, Password: password}
}
// deliverGenesisRegistry seals the genesis document to the build agent on this node as its
// `genesis-package-registry` own-secret (novox/hq ADR 0336 §3), the same way the control plane's store
// connections are delivered — carry the value in, `secret accept`, and the next push writes it sealed
// where the build agent reads it.
//
// **The build agent's one way to the forge until the registry binds.** Its npm-package-registry
// requirement is optional (ADR 0336 §1), and nothing provides it until the gitea module is assigned
// after the base, so the sdk step (15) publishes and the base step (16) resolves with the account in
// this document. It replaces the retired builder's `npm-password` own-secret and the `serves.port`
// setting its carried binding took, neither of which the build agent declares.
func deliverGenesisRegistry(ctx context.Context, o Options, control controlPlane, doc GenesisRegistry,
// deliverBuilderNpm seals the builder's registry password to this node as its `npm-password`
// own-secret, the same way the control plane's store connections are delivered — carry the value in,
// `secret accept`, and the next push writes it sealed where the builder reads it.
func deliverBuilderNpm(ctx context.Context, o Options, control controlPlane, password string,
say func(string)) error {
raw, err := json.Marshal(doc)
if err != nil {
at := "/accepting-npm-password"
if err := control.carryingSecret(ctx, "mesh-accepting-npm-password", []byte(password), at); err != nil {
return err
}
at := "/accepting-" + GenesisRegistrySecret
if err := control.carryingSecret(ctx, "mesh-accepting-"+GenesisRegistrySecret, raw, at); err != nil {
if _, err := control.tell(ctx, "secret", "accept", o.Node, BuilderModule, "npm-password", "--from", at); err != nil {
return err
}
if _, err := control.tell(ctx, "secret", "accept", o.Node, BuilderModule, GenesisRegistrySecret, "--from", at); err != nil {
// Push so the sealed secret reaches the builder, which restarts on it and comes back credentialed.
if _, err := control.tell(ctx, "push", o.Node); err != nil {
return err
}
// Push so the sealed document reaches the build agent, which reads it at its next build — and,
// like every push before the adoption, the bus hears the users it carries (issue 608).
//
// **Waited for until the node applied it** (issue 615). The next step asks the build agent to
// publish the SDK, and the agent reads the document from the file the node writes when it applies
// this push. A push that returned on sending let the ask race the apply: one genesis found the
// document in time, the next was refused, "neither its binding nor the genesis document".
_, err = pushNodeApplied(ctx, o, control, say)
return err
return nil
}
// PublishTheSDK dispatches a build of the SDK to the build agent. The agent has the genesis document by
// now (novox/hq ADR 0336 §7), so the build's `npm publish` authenticates; the artifact is a `package`, built
// PublishTheSDK dispatches a build of the SDK to the builder. The builder has its registry
// credential by now, so the build's `npm publish` authenticates; the artifact is a `package`, built
// on a public base, so this needs no toolchain — which is the whole point of doing it before the base.
func PublishTheSDK(ctx context.Context, o Options, control controlPlane, say func(string)) error {
if o.SDKSource.Repository == "" {
internal/bootstrap/ports.go
+25 -2
View File
@@ -117,6 +117,24 @@ const busModulePort = 4222
// PortsSetting is the controller's settings key for a module's given ports.
const PortsSetting = "ports"
// ServesSetting is the settings key that carries what a provider serves, inside the one binding a
// module carries rather than resolves — the builder's package binding (novox/hq 04-ISSUES/085).
//
// **A port given at genesis has to be a setting wherever it is read, or something puts it back.**
// The builder reaches the package registry through a binding written in its own manifest, because
// at genesis no module yet provides `package-registry` and so there is nothing for the controller
// to resolve it from. That binding used to be rewritten, as text, when the installer registered the
// builder — which a later registration from the catalogue undid silently. Set as a node's setting
// instead, it is held by the controller rather than by the manifest, so re-registering the builder
// leaves it where it was.
//
// **A settings layer is the module's, not one resource's.** The controller lays it over every file
// of that module which merges as JSON, so `serves` lands in the package binding only because that
// binding is the builder's one mergeable file. A second mergeable file added to the builder would
// be given a `serves` key too, meaning nothing to whatever reads it. Worth knowing before adding
// one.
const ServesSetting = "serves"
// setFoundationSettings tells the controller the ports this node gave a foundation module — and,
// on an adopted node, that the registry is reached from anywhere, as a node pulls from it before it
// has a private-network address (novox/hq ADR 0100). Done after the module is registered and before
@@ -130,8 +148,13 @@ func setFoundationSettings(ctx context.Context, o Options, control controlPlane,
if o.Adopted && module == RegistryModule {
values["expose"] = map[string]string{strconv.Itoa(DefaultPorts().Registry): "anywhere"}
}
// No `serves` setting for the build agent: the port given for the package registry reaches it in
// the genesis document (novox/hq ADR 0336 §3), and from the binding once the registry binds.
if packages := o.Ports.orDefaults().Packages; module == BuilderModule &&
packages != DefaultPorts().Packages {
// The one binding nothing resolves: the builder dials the forge by a number it carries.
// The rest of the binding — the provision, the forge it is from, the account it presents —
// is the manifest's and stays there.
values[ServesSetting] = map[string]int{"port": packages}
}
if len(values) == 0 {
return nil
}
internal/bootstrap/ports_test.go
+1 -7
View File
@@ -68,7 +68,7 @@ func TestGivenPortsMoveOnlyTheMachinesSide(t *testing.T) {
if got.Places == 0 {
t.Fatal("nothing was rewritten")
}
storeC := containerNamed(r.Declaration, "postgres")
storeC := containerNamed(r.Declaration, "mesh-store")
if len(storeC.Ports) != 1 || storeC.Ports[0] != "5433:5432" {
t.Errorf("the store publishes %v", storeC.Ports)
}
@@ -239,12 +239,6 @@ func (c *controlRecorder) run(_ context.Context, name string, args ...string) (s
}
if name == "docker" && args[0] == "exec" {
c.told = append(c.told, strings.Join(args[3:], " "))
if strings.Join(args[3:], " ") == "broker accounts" {
return "accounts {}\n", nil // the composed user list, placed after every push (issue 608)
}
}
if name == "sh" {
c.told = append(c.told, "placed the bus's users")
}
return "", nil
}
internal/bootstrap/preflight.go
-7
View File
@@ -127,13 +127,6 @@ func Preflight(ctx context.Context, o Options, d Deps, say func(string)) ([]byte
if err := RefuseAMachineInUse(ctx, o, d.Run, say); err != nil {
return nil, err
}
// The registry's trust is written at step 7, before the push (novox/hq issue 92). What can be
// known about whether it will take, without writing anything, is asked now.
if o.pivots() {
if err := RuntimeWillTakeTheTrust(ctx, o, d.Run, say); err != nil {
return nil, err
}
}
// 4. Can this machine reach what the bundle's images come from?
//
internal/bootstrap/push_places_users_test.go
-134
View File
@@ -1,134 +0,0 @@
package bootstrap
import (
"context"
"strings"
"testing"
"time"
)
// Defends novox/hq issue 608 (ADR 0331 §5, ADR 0333 §4): until the bus is adopted as the nats module,
// every push genesis makes is followed by the installer placing the composed user list where the bus
// reads it, and making the bus re-read it. The build agent's account was minted at step 13 and sealed
// with that push, and the list was placed only at enrolment, so the bus refused the agent
// ("Authorization Violation") and genesis stopped at step 15.
// placing answers the controller as a mesh whose bus is or is not the nats module yet, and records
// what was run, in order.
func placing(adopted bool) (*asked, controlPlane) {
listed := "build-agent 1\n"
if adopted {
listed += "nats 1\n"
}
runtime := &asked{}
runtime.answer = func(name string, args []string) (string, error) {
joined := strings.Join(args, " ")
switch {
case name == "sh":
return "", nil
case strings.HasSuffix(joined, "module list"):
return listed, nil
case strings.HasSuffix(joined, "broker accounts"):
return "accounts {\n MESH { users = [ { user: \"anchor.build-agent\" } ] }\n}\n", nil
}
return "", nil
}
return runtime, controlPlane{container: "mesh-controller", run: runtime.run, timeout: time.Second}
}
// placedAfterThePush is whether the bus's users were placed — written at the module's bus-users path,
// the bus told to re-read them — after the push, and how often.
func placedAfterThePush(t *testing.T, commands []string) int {
t.Helper()
pushed, placed := -1, 0
for i, c := range commands {
if strings.Contains(c, "/mesh-controller push anchor") && pushed < 0 {
pushed = i
}
if strings.HasPrefix(c, "sh -c") && strings.Contains(c, busAccounts) &&
strings.Contains(c, "kill -s HUP "+busContainer) {
if pushed < 0 || i < pushed {
t.Fatalf("the bus's users were placed before the push that carries the account: %v", commands)
}
placed++
}
}
if pushed < 0 {
t.Fatalf("nothing was pushed: %v", commands)
}
return placed
}
func TestTheBusHearsTheBuildAgentsAccountAfterItsPush(t *testing.T) {
t.Setenv("TMPDIR", t.TempDir())
runtime, control := placing(false)
o := Options{Node: "anchor", Wait: time.Second}
if _, err := installModule(context.Background(), o, control, BuilderModule, []byte(`{}`), quietly); err != nil {
t.Fatal(err)
}
if n := placedAfterThePush(t, runtime.commands); n != 1 {
t.Fatalf("the bus's users were placed %d time(s) after the build agent's push; once: %v", n, runtime.commands)
}
}
func TestTheGenesisDocumentsPushIsHeardByTheBusToo(t *testing.T) {
t.Setenv("TMPDIR", t.TempDir())
runtime, control := placing(false)
o := Options{Node: "anchor", Wait: time.Second}
if err := deliverGenesisRegistry(context.Background(), o, control, genesisRegistry(DefaultPorts(), "pw"), quietly); err != nil {
t.Fatal(err)
}
if n := placedAfterThePush(t, runtime.commands); n != 1 {
t.Fatalf("the bus's users were placed %d time(s) after the genesis document's push; once: %v", n, runtime.commands)
}
}
// From the adoption on the nats module writes the list, and the installer does not (ADR 0333 §4).
func TestOnceTheBusIsTheNatsModulesThePushPlacesNothing(t *testing.T) {
t.Setenv("TMPDIR", t.TempDir())
runtime, control := placing(true)
o := Options{Node: "anchor", Wait: time.Second}
if _, err := installModule(context.Background(), o, control, BuilderModule, []byte(`{}`), quietly); err != nil {
t.Fatal(err)
}
if n := placedAfterThePush(t, runtime.commands); n != 0 {
t.Fatalf("the installer placed the bus's users over the nats module's: %v", runtime.commands)
}
}
// Defends novox/hq issue 615: the packages step's push is waited for until the node applied it, because
// the sdk step right after asks the build agent for a build that reads the genesis document from the
// file that apply writes. A push that returned on sending let the ask race the apply.
func TestThePackagesStepWaitsUntilTheNodeAppliedTheGenesisDocument(t *testing.T) {
t.Setenv("TMPDIR", t.TempDir())
runtime, control := placing(false)
o := Options{Node: "anchor", Wait: 3 * time.Minute}
if err := deliverGenesisRegistry(context.Background(), o, control, genesisRegistry(DefaultPorts(), "pw"), quietly); err != nil {
t.Fatal(err)
}
for _, c := range runtime.commands {
if strings.Contains(c, "/mesh-controller push anchor") {
if !strings.HasSuffix(c, "push anchor --wait 3m0s") {
t.Fatalf("the genesis document's push does not wait for the node to apply it: %s", c)
}
return
}
}
t.Fatalf("the genesis document was never pushed: %v", runtime.commands)
}
// Every other push of genesis still returns on sending: only a push whose content the next step reads on
// the machine waits.
func TestAnOrdinaryPushDoesNotWait(t *testing.T) {
t.Setenv("TMPDIR", t.TempDir())
runtime, control := placing(false)
o := Options{Node: "anchor", Wait: time.Second}
if _, err := installModule(context.Background(), o, control, BuilderModule, []byte(`{}`), quietly); err != nil {
t.Fatal(err)
}
for _, c := range runtime.commands {
if strings.Contains(c, "push anchor --wait") {
t.Fatalf("an ordinary push waits: %s", c)
}
}
}
internal/bootstrap/registry_test.go
-12
View File
@@ -58,20 +58,8 @@ func aMeshThatAgrees(answers map[string]string) func(string, []string) (string,
}
}
switch {
case name == "sh" && len(args) == 2 && strings.Contains(args[1], "accounts.conf"):
return "", nil // the bus's users placed after a push (issue 608)
case name == "sh" && len(args) == 2 && strings.Contains(args[1], "mesh-nats-snapshot"):
return "", nil // the streams snapshotted before the bus's planned step (issue 590)
case name == "docker" && strings.HasSuffix(joined, "/mesh-controller bus"):
return "the bus is nats, built 53e51864, on anchor\nlast step 1: not known → 53e51864 on anchor by installer (" +
adoptionWhy + "): done at 2026-10-13 11:40; snapshot " + busSnapshotFile + "\n", nil
case name == "docker" && strings.Contains(joined, "/mesh-controller builds nats -n 1"):
return "nats built 53e51864 anchor\n image artifact-store://nats/server@sha256:" +
strings.Repeat("4", 64) + "\n bundle artifact-store://nats/tools/blobs/sha256:" + strings.Repeat("5", 64) + "\n", nil
case name != "docker":
return "", fmt.Errorf("unexpected program %q", name)
case strings.HasSuffix(joined, "broker accounts"):
return "accounts {}\n", nil
case args[0] == "cp":
return "", nil
case args[0] == "container":
internal/bootstrap/rootsecrets.go
+7 -13
View File
@@ -47,9 +47,6 @@ const (
// StoreSuperuserFile is where the store's superuser password lives on the machine — the
// postgres module's own-secret path, so genesis and adoption write the same file.
StoreSuperuserFile = "/var/lib/postgres/superuser.secret"
// StoreDataDir is where the store keeps its data — the postgres module's data directory, so
// genesis raises the store where the module adopts it (novox/hq ADR 0338 §1).
StoreDataDir = "/var/lib/mesh-store"
// BusControllerFile is where the controller's password on the bus lives on the machine. Not a
// module's path: no module reads it — the permanent controller is handed the address the
// temporary one was given — and genesis keeps it only so a second run puts the same account
@@ -65,7 +62,7 @@ const (
// What the template says, matched exactly. A template that says something else is a template
// this installer does not know how to make safe, and it says so rather than guessing.
templateStorePassword = `"POSTGRES_PASSWORD": "bootstrap"`
templateStoreVolumes = `"volumes": ["` + StoreDataDir + `:/var/lib/postgresql/data"]`
templateStoreVolumes = `"volumes": ["mesh-store-data:/var/lib/postgresql/data"]`
templateStoreURL = "postgres:bootstrap@"
// BusURLPlaceholder is the controller's account in its bus address, as the template names it.
@@ -268,8 +265,8 @@ func Masking(say func(string), c RootCredentials) func(string) {
// RefuseExistingServers stops a run that would put a made credential in front of a store raised
// by an earlier installer with the template's.
//
// The store's password is set by initdb, once, on an empty data directory. A machine that already holds
// a cluster in the store's data directory and has no credential file was raised with `bootstrap`, and minting a new
// The store's password is set by initdb, once, on an empty volume. A machine that already holds
// `mesh-store-data` and has no credential file was raised with `bootstrap`, and minting a new
// value here would make a bundle that dials with a password the server does not have — failing
// three steps later, in the schemas, with nothing pointing back here. Refused by name instead,
// with the way forward.
@@ -281,14 +278,11 @@ func RefuseExistingServers(ctx context.Context, run Runner, c RootCredentials) e
if !c.StoreMade {
return nil
}
// Where the store keeps its data, as the postgres module does (novox/hq ADR 0338 §2): a cluster
// there is one initdb already set a password for. The docker volume genesis used before is not
// the store's data any more, and a machine holding only that is not refused.
if _, err := run(ctx, "test", "-e", StoreDataDir+"/pgdata/PG_VERSION"); err != nil {
return nil // no cluster there: a fresh machine, which is the case this installer makes
if _, err := run(ctx, "docker", "volume", "inspect", "mesh-store-data"); err != nil {
return nil // no such volume: a fresh machine, which is the case this installer makes
}
return fmt.Errorf(
"this machine already holds the store's data (in "+StoreDataDir+") and no credential at %s, so it was raised "+
"this machine already holds the store's data (volume mesh-store-data) and no credential at %s, so it was raised "+
"by an earlier installer with the template's password. A new one made here would not open it. "+
"Put the password the store has into %s (0600, the value alone) and run again; then change it "+
"on the server and accept the new value — this installer does not rotate a running store",
@@ -317,7 +311,7 @@ func RewriteRoot(r *Rewritten, c RootCredentials) (RootRewrite, error) {
return out, err
}
if bundle, err = replaceOnce(bundle, templateStoreVolumes,
`"volumes": ["`+StoreDataDir+`:/var/lib/postgresql/data", "`+StoreSuperuserFile+":"+storeSuperuserMount+`:ro"]`,
`"volumes": ["mesh-store-data:/var/lib/postgresql/data", "`+StoreSuperuserFile+":"+storeSuperuserMount+`:ro"]`,
"the store's volumes"); err != nil {
return out, err
}
internal/bootstrap/rootsecrets_test.go
+1 -1
View File
@@ -59,7 +59,7 @@ func TestTheProducedBundleCarriesNoWellKnownCredential(t *testing.T) {
switch x := res.(type) {
case *declaration.Container:
switch x.Name {
case "postgres":
case "mesh-store":
store = true
if _, has := x.Env["POSTGRES_PASSWORD"]; has {
t.Error("the store still takes its password from its environment")
internal/bootstrap/store_layout_test.go
-163
View File
@@ -1,163 +0,0 @@
package bootstrap
import (
"context"
"encoding/json"
"errors"
"strings"
"testing"
"github.com/novox/mesh-host/internal/declaration"
)
// Defends novox/hq ADR 0338 (issue 609): genesis raises its store as the postgres module's server, so
// the store step adopts it as it runs. Until then the bundle raised `mesh-store`, on another image, in a
// docker volume, and genesis stopped at step 17 of 24.
// theStoreGenesisRaises is the store container of the bundle as the installer applies it: rewritten for
// this machine's credentials, so its password is the file the module reads too.
func theStoreGenesisRaises(t *testing.T) *declaration.Container {
t.Helper()
r := producedBundle(t)
store, err := storeIn(r.Declaration)
if err != nil {
t.Fatal(err)
}
return store
}
// The bundle's store is exactly the catalogue's postgres module's server, read from the module's own
// manifest (ADR 0338 §1): its name, its image (so a drift of either pin fails here), where its data is,
// and the file its superuser password is read from, with the environment that says so.
func TestTheBundleRaisesTheStoreAsTheCataloguesPostgresModuleDeclaresIt(t *testing.T) {
manifest := theCataloguesModule(t, "postgres")
var m struct {
Resources []struct {
ID string `json:"id"`
Type string `json:"type"`
Name string `json:"name"`
Image string `json:"image"`
Env map[string]string `json:"env"`
Ports []string `json:"ports"`
Volumes []string `json:"volumes"`
} `json:"resources"`
}
if err := json.Unmarshal(manifest, &m); err != nil {
t.Fatal(err)
}
ours := theStoreGenesisRaises(t)
found := false
for _, r := range m.Resources {
if r.Type != "container" || r.ID != "server" {
continue
}
found = true
if ours.Name != r.Name {
t.Errorf("the bundle raises the store as %q; the module's server is %q", ours.Name, r.Name)
}
if ours.Image != r.Image {
t.Errorf("the bundle raises the store on %s; the module pins %s — the two pins move together", ours.Image, r.Image)
}
for _, key := range []string{"POSTGRES_PASSWORD_FILE", "PGDATA"} {
if ours.Env[key] != r.Env[key] {
t.Errorf("the bundle's store has %s=%q; the module's %q", key, ours.Env[key], r.Env[key])
}
}
if _, has := ours.Env["POSTGRES_PASSWORD"]; has {
t.Error("the bundle's store takes its password from its environment; the module reads a file")
}
// The module's own directory for its state is <data root>/<module>, where genesis keeps the
// superuser's file (StoreSuperuserFile).
var theirs []string
for _, v := range r.Volumes {
theirs = append(theirs, strings.ReplaceAll(v, "${dir:state}", "/var/lib/postgres"))
}
if strings.Join(ours.Volumes, " ") != strings.Join(theirs, " ") {
t.Errorf("the bundle's store mounts %v; the module's %v", ours.Volumes, theirs)
}
if strings.Join(ours.Ports, " ") != strings.Join(r.Ports, " ") {
t.Errorf("the bundle publishes the store on %v; the module on %v", ours.Ports, r.Ports)
}
// And the store step's own check agrees: by name and image, before anything is registered.
if err := serverMatchesFoundation(manifest, ours, "postgres"); err != nil {
t.Errorf("the store step would refuse the store the bundle raises: %v", err)
}
}
if !found {
t.Fatal("the postgres module declares no server container")
}
}
// The store's data directory is made by the bundle before the store starts, owned by the image's
// postgres user, as the module declares it (ADR 0338 §1, and the measurement under its §5).
func TestTheBundleMakesTheStoresDataDirectoryAsTheModuleDoes(t *testing.T) {
r := producedBundle(t)
at, store := -1, -1
for i, res := range r.Declaration.Resources {
if d, ok := res.(*declaration.Directory); ok && d.Path == StoreDataDir {
at = i
if d.Mode != "0700" || d.Owner != "999:70" {
t.Errorf("the store's data directory is made %s %s; the module makes it 0700 999:70", d.Mode, d.Owner)
}
}
if res.Identity() == StoreID {
store = i
}
}
if at < 0 || store < 0 || at > store {
t.Fatalf("the store's data directory %s is not made before the store (directory at %d, store at %d)", StoreDataDir, at, store)
}
}
// No step of the bundle and no step of the installer names `mesh-store`, which is the seat's name only
// (ADR 0338 §2): every step that reaches the store reaches the container the bundle raised.
func TestNoStepReachesTheStoreByTheSeatsName(t *testing.T) {
text := string(theRealBundle(t))
for _, by := range []string{`"in": "mesh-store"`, `container:mesh-store`, `exec mesh-store`, `"name": "mesh-store"`, `mesh-store-data`} {
if strings.Contains(text, by) {
t.Errorf("the bundle still reaches the store as %s", by)
}
}
store := theStoreGenesisRaises(t)
if foundationStore != store.Name {
t.Errorf("the packages step reaches the store as %q; the bundle raises it as %q", foundationStore, store.Name)
}
r := producedBundle(t)
for _, res := range r.Declaration.Resources {
if a, ok := res.(*declaration.Action); ok && a.In != "" && a.In != store.Name && strings.Contains(a.ID, "database") {
t.Errorf("the step %s reaches the store in %q", a.ID, a.In)
}
}
}
// The installer's refusal to raise a store over data it cannot open looks where the store keeps its data
// (ADR 0338 §2): a cluster there and no credential is refused; the old docker volume alone is not.
func TestTheRefusalOverExistingDataLooksInTheStoresDirectory(t *testing.T) {
made := RootCredentials{Store: "s", Bus: "b", StoreMade: true}
machine := func(cluster, volume bool) Runner {
return func(_ context.Context, name string, args ...string) (string, error) {
joined := name + " " + strings.Join(args, " ")
switch {
case joined == "test -e "+StoreDataDir+"/pgdata/PG_VERSION":
if cluster {
return "", nil
}
case strings.HasPrefix(joined, "docker volume inspect"):
if volume {
return "[]", nil
}
}
return "", errors.New("exit status 1")
}
}
err := RefuseExistingServers(context.Background(), machine(true, false), made)
if err == nil || !strings.Contains(err.Error(), StoreDataDir) {
t.Fatalf("a cluster in %s and no credential was not refused naming it: %v", StoreDataDir, err)
}
if err := RefuseExistingServers(context.Background(), machine(false, true), made); err != nil {
t.Fatalf("a machine holding only the old volume was refused: %v", err)
}
if err := RefuseExistingServers(context.Background(), machine(true, false), RootCredentials{Store: "s"}); err != nil {
t.Fatalf("a kept credential was refused: %v", err)
}
}
internal/bootstrap/trust.go
-317
View File
@@ -1,317 +0,0 @@
package bootstrap
import (
"context"
"encoding/json"
"fmt"
"net"
"os"
"strconv"
"strings"
"time"
"github.com/novox/mesh-host/internal/apply"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
"github.com/novox/mesh-host/internal/system"
)
// The registry's trust, written by genesis before anything is pushed to it (novox/hq issue 92).
//
// The mesh's registry speaks plain HTTP on purpose: every path to it is inside the private
// network's encryption (novox/hq ADR 0082). The container runtime refuses plain HTTP to a registry
// it has not been told to trust, and the module that tells it — `docker`, which writes the
// artifact store's reach into the runtime's configuration — is applied long after genesis pushes
// the control plane's image. So genesis writes that trust itself, first, with the same writer the
// module uses: into the runtime's file beside what is there (ADR 0102), the runtime reloaded and
// never restarted. It then asks the runtime whether it will speak to every name, and refuses with
// the runtime's answer before anything is pushed.
//
// **The handover.** What genesis writes is kept in no record of its own: no record of the foundation
// could hold it, because the retire step re-applies the bundle and takes away a carried record the
// bundle does not name. The `docker` module, applied later, declares the artifact store's reach —
// the mesh name on the node's port — finds it already in the file it has no record of, and takes it
// as its own (ADR 0102: a declared member found in an unrecorded file is the mesh's). That member
// does not change at the handover. The module's first apply usually still changes the file and
// reloads the runtime once, for the keys it writes beside it (`live-restore`, `log-driver`,
// `log-opts`); a reload stops nothing. The address the operator gave, when it is not the mesh name,
// stays as part of the machine's own configuration, as the hand-written trust it replaces would have.
//
// **This rests on an assumption about order.** The module takes the mesh name's member over only
// because its FIRST apply on this node already declares it: genesis assigns no `docker` before the
// network step (16), and once the node is on the network the controller composes the reach as
// `<node>.<suffix>:<port>`.
// Were `docker` applied earlier — while the reach is empty or `127.0.0.1:<port>`, as it is before the
// network — its record would exist without the mesh name, and the member would stay the machine's
// for good: undeclaring the module would never take it out. Installing the node's seat holders
// earlier in genesis must revisit this.
//
// **The suffix** is read from this process's environment, and the controller reads its own: a mesh
// with a custom `MESH_INTERNAL_SUFFIX` sets it for both.
// runtimeConfig is the container runtime's configuration, which the trust is written into. A
// variable so a test writes into a file of its own.
var runtimeConfig = "/etc/docker/daemon.json"
// runtimeUnit is the runtime's service, reloaded when the file changed.
const runtimeUnit = "docker.service"
// internalSuffixVar and defaultInternalSuffix are what a node's mesh name ends in: the same variable
// and default the controller composes `<node>.<suffix>` from (mesh-controller internal/overlay),
// which is the name the artifact store's reach — and so every push by the mesh's builder — uses.
const (
internalSuffixVar = "MESH_INTERNAL_SUFFIX"
defaultInternalSuffix = "internal"
)
// RegistryNames are the names the mesh will use for its registry on this node, each as host:port.
//
// The address the operator gave, and the node's mesh name: the controller's reach for the artifact
// store, which the builder pushes to without being told. Both on the port the registry module
// serves on this node — the node's setting, never the module's default (novox/hq issue 088). A
// loopback name is left out: the runtime trusts loopback without being told, and the mesh never
// names it to another node.
func RegistryNames(o Options) []string {
port := strconv.Itoa(o.Ports.orDefaults().Registry)
var names []string
add := func(hostport string) {
host, _, err := net.SplitHostPort(hostport)
if err != nil || host == "" || isLoopback(host) {
return
}
for _, n := range names {
if n == hostport {
return
}
}
names = append(names, hostport)
}
if host, _, err := net.SplitHostPort(o.Registry); err == nil {
add(net.JoinHostPort(strings.Trim(host, "[]"), port))
}
if node := strings.TrimSpace(o.Node); node != "" {
add(net.JoinHostPort(node+"."+internalSuffix(), port))
}
return names
}
func internalSuffix() string {
if v := strings.TrimSpace(os.Getenv(internalSuffixVar)); v != "" {
return strings.TrimPrefix(v, ".")
}
return defaultInternalSuffix
}
func isLoopback(host string) bool {
if strings.EqualFold(host, "localhost") {
return true
}
ip := net.ParseIP(strings.Trim(host, "[]"))
return ip != nil && ip.IsLoopback()
}
// trustDeclaration is the declaration genesis applies to write the trust: the same two resources
// the `docker` module declares for it — the runtime's file written into, the runtime reloaded on it.
func trustDeclaration(names []string) (*declaration.Declaration, error) {
keys, err := json.Marshal(map[string][]string{"insecure-registries": names})
if err != nil {
return nil, err
}
raw, err := json.Marshal(map[string]any{
"declaration": 1,
"resources": []map[string]any{
{"id": "genesis.registry-trust", "type": "file", "path": runtimeConfig, "mode": "0644",
"into": "json", "content": string(keys) + "\n"},
{"id": "genesis.registry-trust-reload", "type": "service", "unit": runtimeUnit,
"state": "running", "reload-on": []string{"genesis.registry-trust"}},
},
})
if err != nil {
return nil, err
}
return declaration.ParseTrusted(raw)
}
// TrustTheRegistry writes the runtime's trust of every name the mesh serves its registry on here,
// then refuses unless the runtime says it will speak plain HTTP to each.
func TrustTheRegistry(ctx context.Context, o Options, sys system.System, run Runner,
say func(string)) ([]string, error) {
names := RegistryNames(o)
if len(names) == 0 {
say(" trusted only loopback names this registry, which the runtime trusts already")
return nil, nil
}
d, err := trustDeclaration(names)
if err != nil {
return names, err
}
// One apply at a time on this machine, as ApplyBundle does: a host already running here may be
// writing the same file.
unlock, err := store.Lock(o.State, func() { say(" waiting another apply holds this node's state") })
if err != nil {
return names, err
}
report, _, err := apply.Apply(ctx, sys, d, store.State{}, store.OriginCarried, run,
func(line string) { say(" " + strings.TrimPrefix(line, " ")) }, refuseSealed)
unlock()
if err != nil {
return names, fmt.Errorf("the registry's trust could not be written into %s: %w\n"+
"The registry is plain HTTP over the private network (novox/hq ADR 0082), and the container "+
"runtime refuses to push to it until it trusts it. Nothing has been pushed", runtimeConfig, err)
}
if report.Changed() {
say(" trusted " + strings.Join(names, ", ") + " — written into " + runtimeConfig +
", the runtime reloaded")
} else {
say(" trusted " + strings.Join(names, ", ") + " — " + runtimeConfig + " already says so")
}
return names, runtimeWillSpeakTo(ctx, o, run, names)
}
// runtimeWillSpeakTo asks the runtime, until it says it trusts every name or the wait runs out.
//
// The file being right is not the runtime trusting it: a reload the runtime refused, or a flag on
// its command line that contradicts the file, leaves the push to fail at step 8 with an HTTPS error.
func runtimeWillSpeakTo(ctx context.Context, o Options, run Runner, names []string) error {
deadline := time.Now().Add(o.Wait)
var missing []string
var last error
for {
asking, cancel := context.WithTimeout(ctx, o.Timeout)
out, err := run(asking, "docker", "info", "--format", "{{json .RegistryConfig}}")
cancel()
last = err
if err == nil {
missing, last = untrusted(out, names)
if last == nil && len(missing) == 0 {
return nil
}
}
if time.Now().After(deadline) {
break
}
select {
case <-ctx.Done():
return ctx.Err()
case <-time.After(answerEvery):
}
}
if last != nil {
return fmt.Errorf("the container runtime could not be asked whether it trusts the registry: %w", last)
}
return fmt.Errorf(
"the container runtime does not trust %s after %s says it should, so it would refuse to push to "+
"it.\nThe file was written and the runtime reloaded; a runtime that still refuses was started "+
"with its own registry options, or did not take the reload — `journalctl -u %s` says which. "+
"Nothing has been pushed", strings.Join(missing, ", "), runtimeConfig, runtimeUnit)
}
// untrusted reads the runtime's registry configuration and gives back the names it would speak
// HTTPS to: a name is trusted when the runtime lists it as not secure, or when its address falls in
// a range the runtime trusts.
func untrusted(registryConfig string, names []string) ([]string, error) {
var c struct {
InsecureRegistryCIDRs []string `json:"InsecureRegistryCIDRs"`
IndexConfigs map[string]struct {
Secure bool `json:"Secure"`
} `json:"IndexConfigs"`
}
if err := json.Unmarshal([]byte(strings.TrimSpace(registryConfig)), &c); err != nil {
return nil, fmt.Errorf("the runtime's registry configuration is not readable: %w", err)
}
var missing []string
for _, name := range names {
if index, ok := c.IndexConfigs[name]; ok && !index.Secure {
continue
}
host, _, _ := net.SplitHostPort(name)
if ip := net.ParseIP(strings.Trim(host, "[]")); ip != nil && inAny(ip, c.InsecureRegistryCIDRs) {
continue
}
missing = append(missing, name)
}
return missing, nil
}
func inAny(ip net.IP, cidrs []string) bool {
for _, cidr := range cidrs {
if _, n, err := net.ParseCIDR(cidr); err == nil && n.Contains(ip) {
return true
}
}
return false
}
// procRoot is where a process's command line is read from. A variable so a test gives its own.
var procRoot = "/proc"
// RuntimeWillTakeTheTrust is preflight's half of the registry's trust: everything that can be known
// about it without writing anything, asked at step 1 so a machine that cannot take the trust is
// refused before the foundation is raised, rather than at step 7 after it (novox/hq issue 92).
//
// - The runtime's file is absent or a JSON object: the mesh writes into an object, and refuses to
// write over anything else (ADR 0102).
// - The runtime's service is known to systemd and active: the trust is taken by reloading it.
// - The runtime was not started with `--insecure-registry`: dockerd refuses a directive given both
// as a flag and in its file, so it would refuse the reload and keep distrusting the registry.
func RuntimeWillTakeTheTrust(ctx context.Context, o Options, run Runner, say func(string)) error {
names := RegistryNames(o)
if len(names) == 0 {
return nil
}
if raw, err := os.ReadFile(runtimeConfig); err == nil {
var object map[string]json.RawMessage
if len(strings.TrimSpace(string(raw))) > 0 {
if err := json.Unmarshal(raw, &object); err != nil || object == nil {
return fmt.Errorf("%s is not a JSON object, and the mesh writes the trust of its registry "+
"(%s) into it beside what is there, never over it (novox/hq ADR 0102). Make it an object "+
"or remove it; nothing has been changed", runtimeConfig, strings.Join(names, ", "))
}
}
} else if !os.IsNotExist(err) {
return fmt.Errorf("the runtime's configuration %s cannot be read: %w", runtimeConfig, err)
}
asking, cancel := context.WithTimeout(ctx, o.Timeout)
out, err := run(asking, "systemctl", "show", runtimeUnit, "--property=LoadState,ActiveState,MainPID")
cancel()
if err != nil {
return fmt.Errorf("systemd could not be asked about %s: %w\nThe registry's trust is taken by "+
"reloading it; nothing has been changed", runtimeUnit, err)
}
props := map[string]string{}
for _, line := range strings.Split(out, "\n") {
if k, v, ok := strings.Cut(strings.TrimSpace(line), "="); ok {
props[k] = v
}
}
if props["LoadState"] != "loaded" || props["ActiveState"] != "active" {
return fmt.Errorf("%s is %s and %s, and the registry's trust (%s) is taken by reloading it. "+
"Genesis needs the runtime run by systemd as %s; nothing has been changed",
runtimeUnit, orUnknown(props["LoadState"]), orUnknown(props["ActiveState"]),
strings.Join(names, ", "), runtimeUnit)
}
if pid := props["MainPID"]; pid != "" && pid != "0" {
if cmdline, err := os.ReadFile(procRoot + "/" + pid + "/cmdline"); err == nil {
for _, arg := range strings.Split(string(cmdline), "\x00") {
if strings.HasPrefix(arg, "--insecure-registry") {
return fmt.Errorf("the runtime was started with %s, and dockerd refuses insecure registries "+
"given both as a flag and in %s — so it would refuse the reload that makes it trust "+
"%s. Move the flag's registries into the file and restart the runtime once; nothing "+
"has been changed", arg, runtimeConfig, strings.Join(names, ", "))
}
}
}
}
say(" registry trust " + runtimeConfig + " can be written into and " + runtimeUnit + " reloaded")
return nil
}
func orUnknown(s string) string {
if s == "" {
return "unknown"
}
return s
}
internal/bootstrap/trust_test.go
-347
View File
@@ -1,347 +0,0 @@
package bootstrap
import (
"context"
"encoding/json"
"fmt"
"net/http"
"os"
"path/filepath"
"strings"
"testing"
"time"
"github.com/novox/mesh-host/internal/apply"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/store"
"github.com/novox/mesh-host/internal/system"
)
// novox/hq issue 92: genesis pushed the control plane's image to its registry before anything had
// told the container runtime to trust that registry, so the push failed with "server gave HTTP
// response to HTTPS client" — first by the address given, then by the mesh name the builder used,
// and the trust the mesh wrote later named the registry's default port rather than the node's.
// aRuntime is a container runtime on systemd whose registry configuration is what its file said at
// its last reload: what a real dockerd does, and why a written file is not yet a trusted registry.
type aRuntime struct {
path string
trusted []string
commands []string
// ignoresReload is a runtime started with options of its own, which a reload does not change.
ignoresReload bool
}
func (r *aRuntime) run(_ context.Context, name string, args ...string) (string, error) {
line := name + " " + strings.Join(args, " ")
r.commands = append(r.commands, line)
switch {
case name == "docker" && len(args) > 0 && args[0] == "info":
index := map[string]map[string]any{"docker.io": {"Secure": true}}
for _, n := range r.trusted {
index[n] = map[string]any{"Name": n, "Secure": false}
}
out, _ := json.Marshal(map[string]any{"InsecureRegistryCIDRs": []string{"127.0.0.0/8"}, "IndexConfigs": index})
return string(out), nil
case name == "systemctl" && strings.Contains(line, "reload"):
if !r.ignoresReload {
r.trusted = r.trusted[:0]
raw, _ := os.ReadFile(r.path)
var c struct {
Insecure []string `json:"insecure-registries"`
}
_ = json.Unmarshal(raw, &c)
r.trusted = append(r.trusted, c.Insecure...)
}
case strings.Contains(line, "is-enabled"):
return "enabled", nil
case strings.Contains(line, "show") && strings.Contains(line, "ActiveState"):
return "LoadState=loaded\nActiveState=active\nSubState=running", nil
}
return "", nil
}
func (r *aRuntime) did(word string) int {
n := 0
for _, c := range r.commands {
if strings.Contains(c, "systemctl") && strings.Contains(c, " "+word+" ") || strings.HasSuffix(c, " "+word) {
n++
}
}
return n
}
func trustSetup(t *testing.T, existing string) (Options, *aRuntime, system.System) {
t.Helper()
restore := apply.ForTests(t.TempDir(), t.TempDir())
t.Cleanup(restore)
dir := t.TempDir()
was := runtimeConfig
runtimeConfig = filepath.Join(dir, "daemon.json")
t.Cleanup(func() { runtimeConfig = was })
if existing != "" {
if err := os.WriteFile(runtimeConfig, []byte(existing), 0o644); err != nil {
t.Fatal(err)
}
}
t.Setenv(internalSuffixVar, "")
sys, err := system.For("arch")
if err != nil {
t.Fatal(err)
}
o := Options{Node: "anchor", Registry: "192.0.2.10:5100", Ports: FoundationPorts{Registry: 5100},
State: filepath.Join(dir, "state", "state.json"), Timeout: time.Second, Wait: 0}
return o, &aRuntime{path: runtimeConfig}, sys
}
func insecureIn(t *testing.T) (string, map[string]json.RawMessage) {
t.Helper()
raw, err := os.ReadFile(runtimeConfig)
if err != nil {
t.Fatal(err)
}
var o map[string]json.RawMessage
if err := json.Unmarshal(raw, &o); err != nil {
t.Fatal(err)
}
var list []string
_ = json.Unmarshal(o["insecure-registries"], &list)
return fmt.Sprint(list), o
}
func TestGenesisTrustsTheRegistryByEveryNameBeforeItPushes(t *testing.T) {
o, runtime, sys := trustSetup(t, `{"data-root":"/srv/docker","insecure-registries":["198.51.100.7:5000"]}`)
names, err := TrustTheRegistry(context.Background(), o, sys, runtime.run, func(string) {})
if err != nil {
t.Fatal(err)
}
// The address given and the mesh name the builder pushes by, both on the node's port.
if fmt.Sprint(names) != "[192.0.2.10:5100 anchor.internal:5100]" {
t.Errorf("trusted by %v", names)
}
list, object := insecureIn(t)
if list != "[198.51.100.7:5000 192.0.2.10:5100 anchor.internal:5100]" {
t.Errorf("the runtime's list is %s", list)
}
if string(object["data-root"]) != `"/srv/docker"` {
t.Errorf("the machine's own key was not kept: %s", object["data-root"])
}
if runtime.did("reload") != 1 || runtime.did("restart") != 0 || runtime.did("stop") != 0 {
t.Errorf("the runtime was not reloaded exactly once and never restarted: %v", runtime.commands)
}
// Re-run: nothing changes, nothing is reloaded.
runtime.commands = nil
if _, err := TrustTheRegistry(context.Background(), o, sys, runtime.run, func(string) {}); err != nil {
t.Fatal(err)
}
if runtime.did("reload") != 0 {
t.Errorf("a re-run reloaded the runtime: %v", runtime.commands)
}
}
// The port is the node's setting for the registry, not the module's default, whatever the operator
// typed the address with — and loopback, which the runtime trusts already, is not written.
func TestTheRegistrysTrustNamesThePortTheNodeServesItOn(t *testing.T) {
o := Options{Node: "anchor", Registry: "127.0.0.1:5100", Ports: FoundationPorts{Registry: 5100}}
t.Setenv(internalSuffixVar, "")
if got := fmt.Sprint(RegistryNames(o)); got != "[anchor.internal:5100]" {
t.Errorf("names %s", got)
}
o.Ports = FoundationPorts{}
o.Registry = "192.0.2.10:5000"
t.Setenv(internalSuffixVar, ".mesh.example")
if got := fmt.Sprint(RegistryNames(o)); got != "[192.0.2.10:5000 anchor.mesh.example:5000]" {
t.Errorf("names %s", got)
}
}
// A runtime that still refuses after the write is refused here, with the names, before any push.
func TestGenesisRefusesARuntimeThatStillDoesNotTrustTheRegistry(t *testing.T) {
o, runtime, sys := trustSetup(t, "")
runtime.ignoresReload = true
_, err := TrustTheRegistry(context.Background(), o, sys, runtime.run, func(string) {})
if err == nil {
t.Fatal("a runtime that does not trust the registry was let through to the push")
}
if !strings.Contains(err.Error(), "192.0.2.10:5100, anchor.internal:5100") ||
!strings.Contains(err.Error(), "Nothing has been pushed") {
t.Errorf("the refusal does not say what: %v", err)
}
}
// The handover: the `docker` module, applied later, writes the reach the controller composes for
// the artifact store — the mesh name and the node's port. It finds it there: the file does not
// change, the runtime is not reloaded, and the member it declares is recorded as the module's from
// then on (a member found in a file with no record of it is the mesh's, ADR 0102), so the module
// owns the trust it writes and taking the module away takes that member out.
func TestTheDockerModuleTakesTheTrustOverWithoutAChange(t *testing.T) {
o, runtime, sys := trustSetup(t, `{"live-restore":true}`)
if _, err := TrustTheRegistry(context.Background(), o, sys, runtime.run, func(string) {}); err != nil {
t.Fatal(err)
}
before, _ := os.ReadFile(runtimeConfig)
module, err := declaration.ParseTrusted([]byte(fmt.Sprintf(`{"declaration":1,"resources":[
{"id":"docker.daemon","type":"file","path":%q,"mode":"0644","into":"json",
"content":"{\"live-restore\": true, \"insecure-registries\": [\"anchor.internal:5100\"]}\n"},
{"id":"docker.runtime","type":"service","unit":"docker.service","state":"running","boot":"enabled",
"reload-on":["docker.daemon"]}]}`, runtimeConfig)))
if err != nil {
t.Fatal(err)
}
runtime.commands = nil
_, state, err := apply.Apply(context.Background(), sys, module, store.State{}, store.OriginDeclared,
runtime.run, nil, nil)
if err != nil {
t.Fatal(err)
}
after, _ := os.ReadFile(runtimeConfig)
if string(before) != string(after) {
t.Errorf("the handover changed the file:\n%s\n%s", before, after)
}
if runtime.did("reload") != 0 || runtime.did("restart") != 0 {
t.Errorf("the handover touched the runtime: %v", runtime.commands)
}
if rec, _ := state.Find("docker.daemon"); rec.Into == nil || len(rec.Into.Added["insecure-registries"]) != 1 {
t.Errorf("the module did not take the mesh name's member over: %+v", rec.Into)
}
}
func TestARangeTheRuntimeTrustsCoversAnAddressInIt(t *testing.T) {
missing, err := untrusted(`{"InsecureRegistryCIDRs":["10.42.0.0/16"],"IndexConfigs":{}}`,
[]string{"10.42.0.1:5000", "anchor.internal:5000"})
if err != nil {
t.Fatal(err)
}
if fmt.Sprint(missing) != "[anchor.internal:5000]" {
t.Errorf("missing %v", missing)
}
}
// **The order, which is the fault issue 92 is about.** Steps 7 and 8 against a runtime that behaves
// as dockerd does: a push to a registry it has not been told to trust is refused with the error the
// first genesis met. The trust must be written and confirmed before anything is tagged or pushed;
// moving the trust after the push, or dropping it, fails this test.
func TestNothingIsPushedToTheRegistryBeforeTheRuntimeTrustsIt(t *testing.T) {
o, runtime, sys := trustSetup(t, `{"data-root":"/srv/docker"}`)
o.Catalogue = catalogueWith(t, RegistryModule, upstreamRegistryManifest)
o.Registry = "192.0.2.10:5100"
pushed := false
mesh := aMeshThatAgrees(nil)
run := func(ctx context.Context, name string, args ...string) (string, error) {
if name == "systemctl" || (name == "docker" && args[0] == "info") {
return runtime.run(ctx, name, args...)
}
runtime.commands = append(runtime.commands, name+" "+strings.Join(args, " "))
if name == "docker" {
switch args[0] {
case "tag":
return "", nil
case "push":
trusted := false
for _, n := range runtime.trusted {
trusted = trusted || strings.HasPrefix(args[1], n+"/")
}
if !trusted {
return "", fmt.Errorf(`Get "https://%s/v2/": http: server gave HTTP response to HTTPS client`, o.Registry)
}
pushed = true
return "", nil
case "image":
return `["` + o.Registry + `/mesh-controller@sha256:` + strings.Repeat("a", 64) + `"]`, nil
}
}
return mesh(name, args)
}
d := Deps{Run: run, Fetch: func(_ context.Context, url string) (int, string, error) {
if strings.HasSuffix(url, "/v2/") || pushed {
return http.StatusOK, `{"name":"mesh-controller","tags":["genesis"]}`, nil
}
return http.StatusNotFound, "", nil
}}
var result Result
published, err := RegistryAndPublish(context.Background(), o, d, sys,
controlPlane{container: "temp-mesh-controller", run: run, timeout: time.Second}, held, &result, func(string) {})
if err != nil {
t.Fatalf("genesis did not get past the push: %v\ncommands: %v", err, runtime.commands)
}
if !strings.HasPrefix(published.Reference, o.Registry+"/mesh-controller@sha256:") {
t.Errorf("published as %q", published.Reference)
}
firstPush, reloaded, confirmed := -1, -1, -1
for i, c := range runtime.commands {
switch {
case firstPush < 0 && (strings.HasPrefix(c, "docker tag") || strings.HasPrefix(c, "docker push")):
firstPush = i
case strings.HasPrefix(c, "systemctl") && strings.Contains(c, "reload"):
reloaded = i
case strings.HasPrefix(c, "docker info"):
confirmed = i
}
}
if firstPush < 0 || reloaded < 0 || confirmed < 0 || !(reloaded < confirmed && confirmed < firstPush) {
t.Errorf("the trust was not written (%d) and confirmed (%d) before the first tag or push (%d): %v",
reloaded, confirmed, firstPush, runtime.commands)
}
if fmt.Sprint(result.RegistryTrusted) != "[192.0.2.10:5100 anchor.internal:5100]" {
t.Errorf("the result says it trusted %v", result.RegistryTrusted)
}
}
// Preflight asks, without writing, what would stop the trust from taking at step 7.
func TestPreflightRefusesARuntimeThatCouldNotTakeTheTrust(t *testing.T) {
proc := t.TempDir()
was := procRoot
procRoot = proc
t.Cleanup(func() { procRoot = was })
if err := os.MkdirAll(filepath.Join(proc, "42"), 0o755); err != nil {
t.Fatal(err)
}
unit := func(load, active string) Runner {
return func(_ context.Context, name string, args ...string) (string, error) {
return "LoadState=" + load + "\nActiveState=" + active + "\nMainPID=42\n", nil
}
}
cmdline := func(args ...string) {
if err := os.WriteFile(filepath.Join(proc, "42", "cmdline"), []byte(strings.Join(args, "\x00")+"\x00"), 0o644); err != nil {
t.Fatal(err)
}
}
cmdline("/usr/bin/dockerd", "-H", "fd://")
for _, c := range []struct {
name, file string
run Runner
flag bool
refused string
}{
{"a runtime that can take it", `{"data-root":"/srv/docker"}`, unit("loaded", "active"), false, ""},
{"no file yet", "", unit("loaded", "active"), false, ""},
{"a file that is not an object", `["not", "an", "object"]`, unit("loaded", "active"), false, "not a JSON object"},
{"a runtime systemd does not run", `{}`, unit("not-found", "inactive"), false, "not-found and inactive"},
{"a runtime with the flag", `{}`, unit("loaded", "active"), true, "--insecure-registry=10.0.0.1:5000"},
} {
t.Run(c.name, func(t *testing.T) {
o, _, _ := trustSetup(t, c.file)
if c.flag {
cmdline("/usr/bin/dockerd", "--insecure-registry=10.0.0.1:5000")
defer cmdline("/usr/bin/dockerd", "-H", "fd://")
}
before, _ := os.ReadFile(runtimeConfig)
err := RuntimeWillTakeTheTrust(context.Background(), o, c.run, func(string) {})
switch {
case c.refused == "" && err != nil:
t.Errorf("refused: %v", err)
case c.refused != "" && (err == nil || !strings.Contains(err.Error(), c.refused)):
t.Errorf("wanted a refusal naming %q, got %v", c.refused, err)
}
if after, _ := os.ReadFile(runtimeConfig); string(after) != string(before) {
t.Error("preflight wrote the runtime's file")
}
})
}
}
internal/bootstrap/verify_test.go
+1 -1
View File
@@ -89,7 +89,7 @@ func TestAFoundationThatIsUpAndAnsweringIsAccepted(t *testing.T) {
// run-once and scheduled shapes are excluded on purpose — a step that has exited is not a
// fault. The name is `temp-mesh-controller` because the permanent one is a module and takes the
// plain name (novox/hq ADR 0067), which is what makes the two of them coexist at all.
want := []string{"postgres", "mesh-broker-nats", "temp-mesh-controller"}
want := []string{"mesh-store", "mesh-broker-nats", "temp-mesh-controller"}
if len(verified.Running) != len(want) {
t.Fatalf("confirmed %v running, want %v", verified.Running, want)
}
internal/declaration/busy.go
-87
View File
@@ -1,87 +0,0 @@
package declaration
import (
"bytes"
"encoding/json"
"fmt"
"strings"
"time"
)
// Busy is how a long-running container says it holds work that would be lost if it were stopped now
// (novox/hq ADR 0335 §1): a command run inside it, by its shell, as an exec health check is, that exits 0
// while it is busy and non-zero when it is idle. Being asked is also the container's notice that it is
// about to be replaced: from the first ask it takes no new work, so its answer ends once the work in hand
// ends.
//
// **Not part of what the container is.** Busy is not in the container's spec: a changed command changes
// when the container is replaced, never whether — a container is not recreated for its busy command alone.
type Busy struct {
Command string `json:"command"`
// Timeout is how long one ask may take; empty is BusyTimeoutDefault.
Timeout string `json:"timeout,omitempty"`
}
// The bounds of an ask (novox/hq ADR 0335).
const (
BusyTimeoutDefault = 10 * time.Second
// BusyTimeoutCeiling is the longest one ask may take: an ask runs inside an apply, and an apply waits
// for nothing longer than a look.
BusyTimeoutCeiling = time.Minute
// BusyBound is how long a container that says busy is left running after the first apply that would
// have replaced it (ADR 0335 §3). The node-engine's, not the module's: no module holds its own
// replacement for ever.
BusyBound = 15 * time.Minute
)
// UnmarshalJSON reads a busy strictly, as everything a declaration carries is read.
func (b *Busy) UnmarshalJSON(raw []byte) error {
type plain Busy
var p plain
dec := json.NewDecoder(bytes.NewReader(raw))
dec.DisallowUnknownFields()
if err := dec.Decode(&p); err != nil {
return fmt.Errorf("busy: %w", err)
}
*b = Busy(p)
return nil
}
// Within is how long one ask may take. Validated on arrival, so an unreadable timeout is the default.
func (b *Busy) Within() time.Duration {
if d, err := time.ParseDuration(b.Timeout); err == nil && d > 0 {
return d
}
return BusyTimeoutDefault
}
// problems holds a container's busy to its shape: a one-line command, a timeout within its bounds, and only
// on a container that stays up.
func (b *Busy) problems(where string, longRunning bool) []string {
if b == nil {
return nil
}
var problems []string
say := func(format string, args ...any) {
problems = append(problems, where+": "+fmt.Sprintf(format, args...))
}
if !longRunning {
say("busy is asked of a container that stays up; a step or a scheduled run is never replaced mid-run")
}
if strings.TrimSpace(b.Command) == "" {
say("busy needs a command")
}
if strings.ContainsAny(b.Command, "\n\r") {
say("busy's command is one line")
}
if b.Timeout != "" {
d, err := time.ParseDuration(b.Timeout)
switch {
case err != nil:
say("busy's timeout is a duration, not %q", b.Timeout)
case d <= 0 || d > BusyTimeoutCeiling:
say("an ask of busy takes more than nothing and at most %s, not %s", BusyTimeoutCeiling, d)
}
}
return problems
}
internal/declaration/declaration.go
+3 -175
View File
@@ -19,8 +19,6 @@ import (
"sort"
"strconv"
"strings"
"git.novox.be/novox/mesh-sdk/go/seal"
)
// Version is the vocabulary this host speaks. A declaration naming any other version is
@@ -207,12 +205,6 @@ type File struct {
// by looking rather than by knowing which field won.
Sealed string `json:"sealed,omitempty"`
// VaultSealed is content the vault sealed to this node's sealing key, in the mesh's sealed box (novox/hq ADR
// 0206): a core module's own credential, which the controller carries in the declaration and cannot open
// (ADR 0349 §16, issue 755). The same promise as Sealed, from the vault rather than the controller; sent only to
// an engine that states VaultBoxContract. Exclusive with Content, Sealed and Bytes.
VaultSealed *seal.Box `json:"vault-sealed,omitempty"`
// Trusted is the catalogue's word on whether the settings this file's content took are trusted (novox/hq
// issue 339). It is the controller's to act on, and a controller takes it out before it sends a declaration.
// Accepted here, and nothing is done with it, so that a controller that passes it on — an older one, or one
@@ -245,48 +237,11 @@ type File struct {
// Owner is the user this belongs to, by name. Absent means root.
Owner string `json:"owner,omitempty"`
// OwnerAtRuntime says who owns the file once it is placed (novox/hq ADR 0349 §5): "owned", the mesh's
// whole, written back whenever it differs; "seed", written once and then the program's; or "managed",
// only its named keys or its marked block the mesh's, the program keeping the rest.
//
// **Three words for what the forms above already do.** Owned is an ordinary file, seed is create-once
// and managed is a file written into, so the word changes no write: a file that says it is applied
// exactly as its old form is, and one that does not say is owned by its old form (Ownership). The word
// is what the provisioner's files say of themselves, which the old forms only implied. Seed is enough
// alone; managed still needs `into`, which says the format. A word that contradicts the form is refused.
OwnerAtRuntime string `json:"owner-at-runtime,omitempty"`
}
// The words a file says who owns it at runtime with (novox/hq ADR 0349 §5).
const (
RuntimeOwned = "owned"
RuntimeSeed = "seed"
RuntimeManaged = "managed"
)
// Ownership is who owns the file at runtime: the word it says, or the one its old form implies — a file
// written into is managed, one created once is a seed, and any other is owned.
func (f *File) Ownership() string {
switch {
case f.OwnerAtRuntime != "":
return f.OwnerAtRuntime
case f.Into != "":
return RuntimeManaged
case f.CreateOnce:
return RuntimeSeed
default:
return RuntimeOwned
}
}
// Seeded reports whether the file is written only when absent: created once, by create-once or by the
// word seed.
func (f *File) Seeded() bool { return f.CreateOnce || f.OwnerAtRuntime == RuntimeSeed }
// Secret reports whether this file arrived sealed, which is what decides both that it must be
// opened before writing and that its contents must never appear in a report.
func (f *File) Secret() bool { return f.Sealed != "" || f.VaultSealed != nil }
func (f *File) Secret() bool { return f.Sealed != "" }
func (f *File) Identity() string { return f.ID }
func (f *File) Kind() Type { return TypeFile }
@@ -359,13 +314,12 @@ func (f *File) validate(where string, _ bool) []string {
where, f.At))
}
if f.Into == IntoJSON || f.Into == IntoBlock {
if f.Secret() || f.Bytes != "" || len(f.Secrets) > 0 || f.CreateOnce {
if f.Sealed != "" || f.Bytes != "" || len(f.Secrets) > 0 || f.CreateOnce {
problems = append(problems, where+
": a file written into says only its part, in content — not sealed, bytes, "+
"secrets or create-once")
}
}
problems = append(problems, f.checkOwnership(where)...)
var said []string
for name, value := range map[string]string{
"content": f.Content, "sealed": f.Sealed, "bytes": f.Bytes,
@@ -374,16 +328,6 @@ func (f *File) validate(where string, _ bool) []string {
said = append(said, name)
}
}
if f.VaultSealed != nil {
said = append(said, "vault-sealed")
if f.VaultSealed.V != seal.Version {
problems = append(problems, fmt.Sprintf("%s: vault-sealed is a sealed box of version %d, and this "+
"node-engine opens version %d", where, f.VaultSealed.V, seal.Version))
}
if len(f.Secrets) > 0 {
problems = append(problems, where+": a file the vault sealed whole has no holes for secrets")
}
}
if len(said) > 1 {
sort.Strings(said)
problems = append(problems, where+
@@ -417,42 +361,6 @@ func (f *File) validate(where string, _ bool) []string {
return append(problems, checkMode(where, f.Mode)...)
}
// checkOwnership refuses a word that contradicts the file's form (novox/hq ADR 0349 §5): owned is a whole
// file written back, so not created once nor written into; a seed is a whole file created once, so not
// written into; managed is a file written into, which says its format. Managed and created once together
// is already refused by the rule that a file written into is not created once.
func (f *File) checkOwnership(where string) []string {
switch f.OwnerAtRuntime {
case "":
case RuntimeOwned:
var also []string
if f.CreateOnce {
also = append(also, "create-once")
}
if f.Into != "" {
also = append(also, "into")
}
if len(also) > 0 {
return []string{fmt.Sprintf("%s: owned at runtime is the mesh's whole file, written back whenever it "+
"differs, and this also says %s", where, strings.Join(also, " and "))}
}
case RuntimeSeed:
if f.Into != "" {
return []string{where + ": a seed is a whole file written once, and this says into; a file the " +
"program shares with the mesh is managed"}
}
case RuntimeManaged:
if f.Into == "" {
return []string{where + ": a managed file names the format its keys or block are written into, " +
"with into \"json\" or \"block\""}
}
default:
return []string{fmt.Sprintf("%s: owner-at-runtime %q; a placed file is %q, %q or %q, or omits it to be "+
"owned by its form", where, f.OwnerAtRuntime, RuntimeOwned, RuntimeSeed, RuntimeManaged)}
}
return nil
}
// User is a login on the machine.
//
// The thing that makes a shell, a chat client or a desktop expressible at all: each is a package
@@ -1331,11 +1239,6 @@ type Container struct {
// timing, judged by the node-engine beside liveness. Absent: judged alive or not, and nothing more.
Health *Health `json:"health,omitempty"`
// Busy is how this container says it holds work a replacement would lose (novox/hq ADR 0335): asked
// before it is replaced, and while it says busy it is left running, within the bound. Absent: never
// busy, and replaced at once, as before.
Busy *Busy `json:"busy,omitempty"`
// RunAs is who the container runs as, resolved by the controller from the module's identity and the
// image's declared interface (novox/hq ADR 0325 §2, §3): told by environment (the image reads PUID and
// PGID, say) or by the runtime's own user option. Part of the spec, so a change recreates it. Absent: the
@@ -1415,7 +1318,6 @@ func (c *Container) Target() string { return c.Name }
func (c *Container) validate(where string, _ bool) []string {
problems := c.Health.problems(where, true, !c.RunOnce && c.Schedule == "")
problems = append(problems, c.Busy.problems(where, !c.RunOnce && c.Schedule == "")...)
if c.Name == "" {
problems = append(problems, where+": a container needs a name")
}
@@ -1657,71 +1559,6 @@ type Declaration struct {
// this is its home, whether the operator's warrant to take it was carried out, the sub-paths its modules ask for,
// and the tool runner's account to put in its group (§11). Absent is a controller older than ADR 0328.
Libraries []Library
// Filled is the paths of the files the provisioner fills on this machine (novox/hq ADR 0349 §4, §5): each a
// file this declaration declares, which says who owns it at runtime. The provisioner's client has the
// node-engine write only these. Absent is a controller that hands no file to the provisioner, and every
// file is written from the declaration as before.
Filled []string
// Provisioner is the account the provisioner's client runs as on this machine, by name (novox/hq ADR 0349 §4):
// the only account whose write of a filled file the node-engine's local write socket accepts. Absent is a
// controller that names none, and every such write is refused.
Provisioner string
}
// checkProvisioner refuses a provisioner client's account that is no account name, or is root: the account whose
// writes the node-engine accepts must be one that could not write the file without it.
func checkProvisioner(name string) []string {
switch {
case name == "":
return nil
case !accountName.MatchString(name):
return []string{fmt.Sprintf("provisioner_account %q is not an account name", name)}
case name == "root":
return []string{"provisioner_account is root; the provisioner's client runs as an account of its own"}
}
return nil
}
// Fills reports whether the provisioner fills the file at path on this machine.
func (d *Declaration) Fills(path string) bool { return slices.Contains(d.Filled, path) }
// checkFilled refuses a filled path that is not a declared file saying who owns it at runtime, or that is named
// twice or empty. Every file declared at the path must say it: the provisioner writes the path, and a file there
// the mesh writes some other way would be a second writer of it (ADR 0349 §5).
func checkFilled(filled []string, resources []Resource) []string {
var problems []string
files := map[string][]*File{}
for _, r := range resources {
if f, ok := r.(*File); ok {
files[f.Path] = append(files[f.Path], f)
}
}
seen := map[string]bool{}
for _, path := range filled {
switch {
case path == "":
problems = append(problems, "filled_files names a file with no path")
continue
case seen[path]:
problems = append(problems, fmt.Sprintf("filled_files names %q twice", path))
continue
}
seen[path] = true
if len(files[path]) == 0 {
problems = append(problems, fmt.Sprintf("filled_files names %q, and this declaration declares no "+
"file there", path))
continue
}
for _, f := range files[path] {
if f.OwnerAtRuntime == "" {
problems = append(problems, fmt.Sprintf("resource %q: the provisioner fills %q, and a file it "+
"fills says who owns it at runtime with owner-at-runtime", f.ID, path))
}
}
}
return problems
}
// Identity is one module's identity, as the controller allocated it (ADR 0328 §4): its user `mesh-<module>` with
@@ -2061,12 +1898,6 @@ type envelope struct {
// sends them only to a host whose health says link.LibraryContract.
Groups []Group `json:"groups,omitempty"`
Libraries []Library `json:"libraries,omitempty"`
// Filled is optional too (novox/hq ADR 0349 §5), as is a file's `owner-at-runtime`. **An older host refuses
// these keys**; a controller sends them only to a host whose health says link.OwnershipContract.
Filled []string `json:"filled_files,omitempty"`
// Provisioner is optional too (novox/hq ADR 0349 §4). **An older host refuses this key**; a controller sends it
// only to a host whose health says link.ProvisionerContract.
Provisioner string `json:"provisioner_account,omitempty"`
}
func parse(raw []byte, allowActions bool) (*Declaration, error) {
@@ -2085,10 +1916,8 @@ func parse(raw []byte, allowActions bool) (*Declaration, error) {
d := &Declaration{Version: env.Version, For: env.For, Adoption: env.Adoption, Sequence: env.Sequence,
Epoch: env.Epoch, LeftOut: env.LeftOut, Generation: env.Generation, PutBack: env.PutBack,
Identities: env.Identities, Groups: env.Groups, Libraries: env.Libraries, Filled: env.Filled,
Provisioner: env.Provisioner}
Identities: env.Identities, Groups: env.Groups, Libraries: env.Libraries}
problems := checkIdentities(env.Identities, env.Groups)
problems = append(problems, checkProvisioner(env.Provisioner)...)
problems = append(problems, checkLibraries(env.Groups, env.Libraries)...)
if env.Generation < 0 {
// Below zero would read as "none claimed" and pass every refusal of an older generation.
@@ -2180,7 +2009,6 @@ func parse(raw []byte, allowActions bool) (*Declaration, error) {
problems = append(problems, checkWhileStopped(d.Resources)...)
problems = append(problems, checkAdoption(env.Adoption, d.Resources, allowActions)...)
problems = append(problems, checkReplaces(d.Resources)...)
problems = append(problems, checkFilled(env.Filled, d.Resources)...)
if env.Adoption == nil {
for _, r := range d.Resources {
if r.Kind() == TypeOpening {
internal/declaration/ownership_test.go
-130
View File
@@ -1,130 +0,0 @@
package declaration
import (
"strings"
"testing"
)
// Defends novox/hq ADR 0349 §5 (issue 667): a file the provisioner places says who owns it at runtime
// — owned, seed or managed — and the declaration names the files the provisioner fills. The words are
// read and judged here before any controller sends them.
func TestAFileSaysWhoOwnsItAtRuntime(t *testing.T) {
d, err := Parse([]byte(`{"declaration":1,"resources":[
{"id":"o","type":"file","path":"/etc/o","content":"a","owner-at-runtime":"owned"},
{"id":"s","type":"file","path":"/etc/s","content":"a","owner-at-runtime":"seed"},
{"id":"s2","type":"file","path":"/etc/s2","content":"a","owner-at-runtime":"seed","create-once":true},
{"id":"mj","type":"file","path":"/etc/mj","into":"json","content":"{\"k\":1}","owner-at-runtime":"managed"},
{"id":"mb","type":"file","path":"/etc/mb","into":"block","content":"a\n","owner-at-runtime":"managed"}
]}`))
if err != nil {
t.Fatal(err)
}
want := map[string]string{"o": RuntimeOwned, "s": RuntimeSeed, "s2": RuntimeSeed, "mj": RuntimeManaged,
"mb": RuntimeManaged}
for _, r := range d.Resources {
f := r.(*File)
if f.OwnerAtRuntime != want[f.ID] || f.Ownership() != want[f.ID] {
t.Errorf("%s: read as %q, ownership %q; want %q", f.ID, f.OwnerAtRuntime, f.Ownership(), want[f.ID])
}
}
}
// A file that does not say is owned by its old form, so the word means exactly what that form does.
func TestAFileThatDoesNotSayIsOwnedByItsOldForm(t *testing.T) {
d, err := Parse([]byte(`{"declaration":1,"resources":[
{"id":"o","type":"file","path":"/etc/o","content":"a"},
{"id":"s","type":"file","path":"/etc/s","content":"a","create-once":true},
{"id":"mj","type":"file","path":"/etc/mj","into":"json","content":"{}"},
{"id":"mb","type":"file","path":"/etc/mb","into":"block","content":"a\n"}
]}`))
if err != nil {
t.Fatal(err)
}
want := map[string]string{"o": RuntimeOwned, "s": RuntimeSeed, "mj": RuntimeManaged, "mb": RuntimeManaged}
for _, r := range d.Resources {
f := r.(*File)
if f.OwnerAtRuntime != "" || f.Ownership() != want[f.ID] {
t.Errorf("%s: said %q, ownership %q; want unsaid and %q", f.ID, f.OwnerAtRuntime, f.Ownership(), want[f.ID])
}
if f.Seeded() != (want[f.ID] == RuntimeSeed) {
t.Errorf("%s: seeded %v", f.ID, f.Seeded())
}
}
}
func TestAnOwnershipThatContradictsTheFileIsRefused(t *testing.T) {
for name, c := range map[string]struct{ resource, refusal string }{
"another word": {`{"id":"f","type":"file","path":"/etc/x","content":"a","owner-at-runtime":"mine"}`,
`owner-at-runtime "mine"`},
"owned and create-once": {`{"id":"f","type":"file","path":"/etc/x","content":"a","owner-at-runtime":"owned","create-once":true}`,
"create-once"},
"owned and into": {`{"id":"f","type":"file","path":"/etc/x","into":"json","content":"{}","owner-at-runtime":"owned"}`,
"into"},
"seed and into": {`{"id":"f","type":"file","path":"/etc/x","into":"block","content":"a","owner-at-runtime":"seed"}`,
"into"},
"managed and whole": {`{"id":"f","type":"file","path":"/etc/x","content":"a","owner-at-runtime":"managed"}`,
"into"},
"managed and create-once": {`{"id":"f","type":"file","path":"/etc/x","into":"json","content":"{}","owner-at-runtime":"managed","create-once":true}`,
"create-once"},
// The rule that was there before stays: a file written into is never sealed or created once.
"managed and sealed": {`{"id":"f","type":"file","path":"/etc/x","into":"json","content":"{}","sealed":"x","owner-at-runtime":"managed"}`,
"sealed"},
} {
_, err := Parse([]byte(`{"declaration":1,"resources":[` + c.resource + `]}`))
if err == nil || !strings.Contains(err.Error(), c.refusal) {
t.Errorf("%s: want a refusal naming %q, got %v", name, c.refusal, err)
}
}
}
func TestTheDeclarationNamesTheFilesTheProvisionerFills(t *testing.T) {
d, err := Parse([]byte(`{"declaration":1,"filled_files":["/etc/a","/etc/b"],"resources":[
{"id":"a","type":"file","path":"/etc/a","content":"a","owner-at-runtime":"owned"},
{"id":"b","type":"file","path":"/etc/b","into":"block","content":"b\n","owner-at-runtime":"managed"},
{"id":"c","type":"file","path":"/etc/c","content":"c"}
]}`))
if err != nil {
t.Fatal(err)
}
if len(d.Filled) != 2 || d.Filled[0] != "/etc/a" || d.Filled[1] != "/etc/b" {
t.Fatalf("filled files read as %v", d.Filled)
}
if !d.Fills("/etc/a") || !d.Fills("/etc/b") || d.Fills("/etc/c") {
t.Errorf("Fills: a %v, b %v, c %v", d.Fills("/etc/a"), d.Fills("/etc/b"), d.Fills("/etc/c"))
}
// A declaration that names none fills none, as every declaration before this did.
d, err = Parse([]byte(`{"declaration":1,"resources":[{"id":"c","type":"file","path":"/etc/c","content":"c"}]}`))
if err != nil {
t.Fatal(err)
}
if d.Filled != nil || d.Fills("/etc/c") {
t.Errorf("a declaration naming no filled files fills %v", d.Filled)
}
}
func TestAFilledFileMustBeADeclaredFileThatSaysItsOwner(t *testing.T) {
for name, c := range map[string]struct{ filled, resources, refusal string }{
"not declared": {`["/etc/none"]`,
`{"id":"a","type":"file","path":"/etc/a","content":"a","owner-at-runtime":"owned"}`,
`"/etc/none"`},
"a directory, not a file": {`["/etc/d"]`,
`{"id":"d","type":"directory","path":"/etc/d"}`,
`"/etc/d"`},
"says no owner": {`["/etc/a"]`,
`{"id":"a","type":"file","path":"/etc/a","content":"a"}`,
"owner-at-runtime"},
"named twice": {`["/etc/a","/etc/a"]`,
`{"id":"a","type":"file","path":"/etc/a","content":"a","owner-at-runtime":"owned"}`,
"twice"},
"no path": {`[""]`,
`{"id":"a","type":"file","path":"/etc/a","content":"a","owner-at-runtime":"owned"}`,
"no path"},
} {
_, err := Parse([]byte(`{"declaration":1,"filled_files":` + c.filled + `,"resources":[` + c.resources + `]}`))
if err == nil || !strings.Contains(err.Error(), c.refusal) {
t.Errorf("%s: want a refusal naming %q, got %v", name, c.refusal, err)
}
}
}
internal/declaration/provisioner_test.go
-37
View File
@@ -1,37 +0,0 @@
package declaration
import (
"strings"
"testing"
)
// Defends novox/hq ADR 0349 §4 (issue 668): the declaration names the account the provisioner's client runs as,
// the only one whose write of a filled file the node-engine accepts. Read here before any controller sends it.
func TestTheDeclarationNamesTheProvisionersAccount(t *testing.T) {
d, err := Parse([]byte(`{"declaration":1,"provisioner_account":"mesh-node-provisioner","filled_files":["/etc/a"],
"resources":[{"id":"a","type":"file","path":"/etc/a","content":"a","owner-at-runtime":"owned"}]}`))
if err != nil {
t.Fatal(err)
}
if d.Provisioner != "mesh-node-provisioner" {
t.Errorf("the provisioner's account read as %q", d.Provisioner)
}
d, err = Parse([]byte(`{"declaration":1,"resources":[{"id":"a","type":"file","path":"/etc/a","content":"a"}]}`))
if err != nil {
t.Fatal(err)
}
if d.Provisioner != "" {
t.Errorf("a declaration naming no account names %q", d.Provisioner)
}
}
func TestAProvisionersAccountThatIsNoAccountIsRefused(t *testing.T) {
for name, account := range map[string]string{"root": "root", "not a name": "a b", "a path": "../x"} {
_, err := Parse([]byte(`{"declaration":1,"provisioner_account":"` + account + `",
"resources":[{"id":"a","type":"file","path":"/etc/a","content":"a"}]}`))
if err == nil || !strings.Contains(err.Error(), "provisioner_account") {
t.Errorf("%s: want a refusal naming provisioner_account, got %v", name, err)
}
}
}
internal/filling/filling.go
-512
View File
@@ -1,512 +0,0 @@
// Package filling is the node-engine's narrow local write verb (novox/hq ADR 0349 §4, design 57 P2): a socket on
// this machine through which the provisioner's client has a file it filled written, and nothing else.
//
// **The client fills; the node-engine writes.** The client renders a module's template with the values each domain
// owns and hands the node-engine the content; the node-engine writes it as the declared file, with the declared
// file's path, owner, mode and ownership at runtime, and makes what reads it take the change. So the one process that
// writes the machine's files is still the node-engine, and the client needs no right to write any of them.
//
// **Narrow by three checks, each from something the client cannot say for itself.** The account asking, from the
// kernel (SO_PEERCRED), in this machine's own namespaces, must be the account the node's current declaration names
// as the provisioner's client's; the path, clean and absolute, must be one the declaration names as filled; and the
// write takes the apply's lock, so a write and an apply never write one file at once. While the declaration names no
// account — every declaration before the provisioner's client has one — every write is refused.
//
// **No listening control surface** (novox/hq ADR 0004): the socket has no network address. And the content is never
// said: the journal has the path, the account and the content's digest.
package filling
import (
"bytes"
"context"
"encoding/json"
"errors"
"fmt"
"io"
"net"
"os"
"os/user"
"path/filepath"
"sort"
"strconv"
"sync"
"sync/atomic"
"time"
"github.com/novox/mesh-host/internal/declaration"
"github.com/novox/mesh-host/internal/meshcli"
)
// Socket is where the node-engine accepts a write of a filled file, beside mesh-cli's socket.
const Socket = "/run/nox-mesh-host/fill.sock"
// MaxRequest bounds one request, its content encoded in it: a configuration file, never an archive.
const MaxRequest = 8 << 20
// Request is one write, sent as one JSON object on one line: the path of a filled file and what the client filled
// into it (base64 in the JSON). The content only: the declaration says everything else about the file.
type Request struct {
Path string `json:"path"`
Content []byte `json:"content"`
}
// Answer is what came of one request. Refused is a write that was not attempted, with why; Failed one that was
// attempted and did not complete, with why — the file, or something that reads it. Otherwise Action is what the
// write did to the file (created, updated, corrected, unchanged, kept) and Digest what the file holds now, as the
// node-engine records it.
type Answer struct {
Path string `json:"path,omitempty"`
Action string `json:"action,omitempty"`
Digest string `json:"digest,omitempty"`
Refused string `json:"refused,omitempty"`
Failed string `json:"failed,omitempty"`
}
// Refusal is an answer that nothing was written, and why.
func Refusal(path, why string) Answer { return Answer{Path: path, Refused: why} }
// Writer is the node-engine's side of the socket.
type Writer interface {
// Accepts judges the asking account before its request is read: nil when the current declaration names it as
// the provisioner's client's.
Accepts(uid uint32) error
// Write judges the request and writes it, holding the apply's lock.
Write(ctx context.Context, uid uint32, req Request) Answer
}
// Engine is the Writer the node-engine serves: the declaration it applies now, read again under the apply's lock.
type Engine struct {
// Declared reads the declaration this node applies now, verified.
Declared func() (*declaration.Declaration, error)
// UIDOf is the user id of an account, by name, from this machine's user database.
UIDOf func(account string) (uint32, error)
// Lock takes the apply's lock and returns what releases it: the same the apply takes, so a write and an apply
// are one at a time.
Lock func() (func(), error)
// Place writes the content into the filled file at path, as the declared file, and makes what reads it take the
// change. Called holding the lock, with the declaration read under it.
Place func(ctx context.Context, d *declaration.Declaration, path string, content []byte) (action, digest string,
err error)
}
// Accepts reads the declaration and judges the account.
func (e *Engine) Accepts(uid uint32) error {
d, err := e.Declared()
if err != nil {
return fmt.Errorf("this node's declaration cannot be read, so no write is accepted: %w", err)
}
return JudgeAccount(d, uid, e.UIDOf)
}
// Write takes the apply's lock, reads the declaration under it, judges the account and the path again — the
// declaration may have changed since the request was accepted — and writes.
func (e *Engine) Write(ctx context.Context, uid uint32, req Request) Answer {
if err := JudgeShape(req.Path); err != nil {
return Refusal(req.Path, err.Error())
}
unlock, err := e.Lock()
if err != nil {
return Refusal(req.Path, "the apply's lock could not be taken, so nothing was written: "+err.Error())
}
defer unlock()
d, err := e.Declared()
if err != nil {
return Refusal(req.Path, "this node's declaration cannot be read, so nothing was written: "+err.Error())
}
if err := JudgeAccount(d, uid, e.UIDOf); err != nil {
return Refusal(req.Path, err.Error())
}
if err := JudgePath(d, req.Path); err != nil {
return Refusal(req.Path, err.Error())
}
action, digest, err := e.Place(ctx, d, req.Path, req.Content)
if err != nil {
return Answer{Path: req.Path, Action: action, Digest: digest, Failed: err.Error()}
}
return Answer{Path: req.Path, Action: action, Digest: digest}
}
// JudgeAccount refuses every account but the one the declaration names as the provisioner's client's, by its user
// id here — and every account while the declaration names none.
func JudgeAccount(d *declaration.Declaration, uid uint32, uidOf func(string) (uint32, error)) error {
if d.Provisioner == "" {
return errors.New("this node's declaration names no provisioner client's account, so no write is accepted")
}
want, err := uidOf(d.Provisioner)
if err != nil {
return fmt.Errorf("the provisioner client's account %s is not on this machine, so no write is accepted: %w",
d.Provisioner, err)
}
if want == 0 {
return fmt.Errorf("the provisioner client's account %s is root here, so no write is accepted", d.Provisioner)
}
if uid != want {
return fmt.Errorf("user id %d is not the provisioner client's account %s (user id %d), so nothing was written",
uid, d.Provisioner, want)
}
return nil
}
// JudgeShape refuses a path that is not clean and absolute: a relative path, `..`, a doubled or trailing slash.
// Compared as written with the declaration's, a path that is not clean could name a file the declaration does not.
func JudgeShape(path string) error {
switch {
case path == "":
return errors.New("the request names no path, so nothing was written")
case !filepath.IsAbs(path):
return fmt.Errorf("%q is not an absolute path, so nothing was written", path)
case filepath.Clean(path) != path:
return fmt.Errorf("%q is not a clean path (it is %q cleaned), so nothing was written", path,
filepath.Clean(path))
}
return nil
}
// JudgePath refuses a path that is not clean and absolute, or that the declaration does not name as filled. Where it
// leads — a link on the way, a place nothing is written — is judged by the write, by the apply's own rules.
func JudgePath(d *declaration.Declaration, path string) error {
if err := JudgeShape(path); err != nil {
return err
}
if !d.Fills(path) {
return fmt.Errorf("%s is not a file this node's declaration names as filled, so nothing was written", path)
}
return nil
}
// UIDOf is the user id of an account by name, from this machine's user database.
func UIDOf(account string) (uint32, error) {
u, err := user.Lookup(account)
if err != nil {
return 0, err
}
id, err := strconv.ParseUint(u.Uid, 10, 32)
if err != nil {
return 0, fmt.Errorf("user id %q: %w", u.Uid, err)
}
return uint32(id), nil
}
// peerOf and sameNamespaces are the kernel's word on the asking process, as mesh-cli's socket reads it. Variables so
// a test can say.
var (
peerOf = meshcli.PeerCredentials
sameNamespaces = meshcli.SameNamespaces
)
// within bounds how long an accepted connection may take to say its request, and refusedWithin how long a busy one
// is read and dropped before its refusal is written (so the client is not answered with a broken pipe). Variables so
// a test need not wait.
var (
within = 30 * time.Second
refusedWithin = 1 * time.Second
)
// atOnce bounds the accepted connections served at the same time: one client, writing one file at a time. One over
// is refused busy. busyWaitingAtOnce bounds how many busy refusals wait for their request at once, per Serve; over it
// a busy refusal is written at once, as mesh-cli's socket does. judgingAtOnce bounds the callers judged at once, and
// judgeWaiting the connections waiting to be judged — each an open file, so far below the engine's open-file limit;
// one over is closed at once, unread. Variables so a test can say.
var (
atOnce = 4
busyWaitingAtOnce = 8
judgingAtOnce = 16
judgeWaiting = 256
)
// refusalsSaidEvery is how often the refusals counted and not yet said are said: each kind of refusal of each account
// is said at once the first time, and what follows it within this time is counted and said at its end, or when the
// socket stops. A variable so a test need not wait.
var refusalsSaidEvery = time.Minute
// acceptBackoff is how long the socket waits after an accept that failed before it accepts again, at first and at
// most: a socket that stopped at the first such failure (too many open files, say) would stop the client's writes
// until the engine restarted. Variables so a test need not wait.
var (
acceptBackoff = 50 * time.Millisecond
acceptBackoffMax = 5 * time.Second
)
// acceptFrom accepts one connection. A variable so a test can make it fail.
var acceptFrom = func(l *net.UnixListener) (*net.UnixConn, error) { return l.AcceptUnix() }
// Serve listens on socket until ctx ends, and answers each connection: one request, one answer.
//
// The directory is made 0755 and the socket 0666, as mesh-cli's: who may write is judged per connection from the
// kernel's word, against the current declaration, so the socket needs no account that may not exist yet. **The
// caller is judged before it holds anything**: one the kernel does not say is the provisioner client's account is
// answered and closed at once, and only the client's connections are served or wait as busy, so no other account can
// hold the socket from it. A file at the socket's path that is not a socket is never removed: the engine says so and
// serves nothing.
func Serve(ctx context.Context, socket string, w Writer, say func(string)) error {
if say == nil {
say = func(string) {}
}
dir := filepath.Dir(socket)
if err := os.MkdirAll(dir, 0o755); err != nil {
return fmt.Errorf("the filled files' socket directory: %w", err)
}
if problem := meshcli.NotOwnDirectory(dir); problem != "" {
return fmt.Errorf("filled files are not written on request on this machine: %s", problem)
}
if err := os.Chmod(dir, 0o755); err != nil {
return fmt.Errorf("the filled files' socket directory: %w", err)
}
if info, err := os.Lstat(socket); err == nil {
if info.Mode()&os.ModeSocket == 0 {
return fmt.Errorf("%s is there and is not a socket, and the node-engine did not put it there; filled "+
"files are not written on request until it is moved", socket)
}
if err := os.Remove(socket); err != nil {
return fmt.Errorf("the socket a previous engine left at %s: %w", socket, err)
}
}
l, err := net.ListenUnix("unix", &net.UnixAddr{Name: socket, Net: "unix"})
if err != nil {
return fmt.Errorf("listening for filled files at %s: %w", socket, err)
}
if err := os.Chmod(socket, 0o666); err != nil {
l.Close()
return fmt.Errorf("opening the filled files' socket: %w", err)
}
go func() {
<-ctx.Done()
l.Close()
}()
slots := make(chan struct{}, atOnce)
// Judging reads the declaration: at most judging callers are judged at once, and the rest wait their turn — a
// few milliseconds each — rather than being refused, so a flood delays the client's write and never stops it.
judging := make(chan struct{}, judgingAtOnce)
var waiting atomic.Int64
busyWaiting := make(chan struct{}, busyWaitingAtOnce)
refusals := &refusalWords{say: say}
// Said once everything has stopped: what was counted is never lost with the socket.
defer refusals.flush()
var wg sync.WaitGroup
defer wg.Wait()
wg.Add(1)
go func() {
defer wg.Done()
tick := time.NewTicker(refusalsSaidEvery)
defer tick.Stop()
for {
select {
case <-ctx.Done():
return
case <-tick.C:
refusals.flush()
}
}
}()
backoff := time.Duration(0)
for {
conn, err := acceptFrom(l)
if err != nil {
if ctx.Err() != nil {
return nil
}
if errors.Is(err, net.ErrClosed) {
return fmt.Errorf("the filled files' socket was closed: %w", err)
}
// Said, waited for, and accepted again: the failure is the machine's for a moment, not the socket's.
backoff = min(max(2*backoff, acceptBackoff), acceptBackoffMax)
say(fmt.Sprintf("the filled files' socket could not accept a connection (%v); accepting again in %s",
err, backoff))
select {
case <-ctx.Done():
return nil
case <-time.After(backoff):
}
continue
}
backoff = 0
if waiting.Add(1) > int64(judgeWaiting) {
// Too many waiting to be judged already: this one is closed unread, and counted.
waiting.Add(-1)
refusals.said(^uint32(0), "overflow", "too many connections are waiting to be judged; closed unread")
conn.Close()
continue
}
wg.Add(1)
go func() {
defer wg.Done()
// Judged before it holds a slot or is waited for.
judging <- struct{}{}
uid, ok := judgePeer(conn, w, refusals)
<-judging
waiting.Add(-1)
if !ok {
return
}
readBy := time.Now().Add(within)
_ = conn.SetDeadline(readBy)
select {
case slots <- struct{}{}:
defer func() { <-slots }()
answer(ctx, conn, uid, w, say)
default:
busy(conn, readBy, busyWaiting, say)
}
}()
}
}
// judgePeer reads the asking account from the kernel and judges it. A caller that is not the provisioner client's
// account is answered with why and closed at once, its request never read: false.
func judgePeer(conn *net.UnixConn, w Writer, refusals *refusalWords) (uint32, bool) {
closeWith := func(uid uint32, kind, why string) (uint32, bool) {
refusals.said(uid, kind, why)
_ = conn.SetWriteDeadline(time.Now().Add(refusedWithin))
_ = json.NewEncoder(conn).Encode(Refusal("", why))
conn.Close()
return uid, false
}
uid, pid, err := peerOf(conn)
if err != nil {
return closeWith(^uint32(0), "unread", "the node-engine could not read which account is asking, so nothing was "+
"written: "+err.Error())
}
if err := sameNamespaces(pid); err != nil {
return closeWith(uid, "namespace", "the asking process does not run in this machine's own namespaces, so nothing was "+
"written: "+err.Error())
}
if err := w.Accepts(uid); err != nil {
return closeWith(uid, "account", err.Error())
}
return uid, true
}
// refusalWords says refused callers in the journal without letting a flood flood it: the first refusal of each kind
// for each account at once, and those that follow it counted and said by flush — on a timer and when the socket stops
// — so an account that stops still has its count said, and a refusal of another kind never hides behind one said.
type refusalWords struct {
mu sync.Mutex
say func(string)
pending map[refusalKey]*refusalCount
}
type refusalKey struct {
uid uint32
kind string
}
type refusalCount struct {
n int
why string
}
func (r *refusalWords) said(uid uint32, kind, why string) {
r.mu.Lock()
defer r.mu.Unlock()
if r.pending == nil {
r.pending = map[refusalKey]*refusalCount{}
}
k := refusalKey{uid, kind}
if c, ok := r.pending[k]; ok {
c.n++
c.why = why
return
}
r.pending[k] = &refusalCount{}
r.say(fmt.Sprintf("a filled file's write from %s was refused (%s): %s", who(k), kind, why))
}
// flush says what was counted since each was last said, and forgets every key, so the next refusal of each is said at
// once again.
func (r *refusalWords) flush() {
r.mu.Lock()
defer r.mu.Unlock()
keys := make([]refusalKey, 0, len(r.pending))
for k := range r.pending {
keys = append(keys, k)
}
sort.Slice(keys, func(i, j int) bool {
if keys[i].uid != keys[j].uid {
return keys[i].uid < keys[j].uid
}
return keys[i].kind < keys[j].kind
})
for _, k := range keys {
if c := r.pending[k]; c.n > 0 {
r.say(fmt.Sprintf("%d more writes of a filled file from %s were refused (%s) since it was last said; the last: %s",
c.n, who(k), k.kind, c.why))
}
}
r.pending = nil
}
// who names a caller by its user id, or says why it has none.
func who(k refusalKey) string {
switch {
case k.kind == "overflow":
return "a caller not yet judged"
case k.uid == ^uint32(0):
return "an account that could not be read"
}
return fmt.Sprintf("user id %d", k.uid)
}
// answer serves one connection from the provisioner client's account.
func answer(ctx context.Context, conn *net.UnixConn, uid uint32, w Writer, say func(string)) {
var req Request
dec := json.NewDecoder(io.LimitReader(conn, MaxRequest))
dec.DisallowUnknownFields()
if err := dec.Decode(&req); err != nil {
// Answered even when the deadline to say it passed: the deadline covered the write too.
_ = conn.SetWriteDeadline(time.Now().Add(refusedWithin))
reply(conn, Refusal("", "the request could not be read, so nothing was written: "+err.Error()), say)
return
}
// The write waits for the apply's lock as long as an apply takes, and the client waits with it.
_ = conn.SetDeadline(time.Time{})
a := w.Write(ctx, uid, req)
switch {
case a.Refused != "":
say(fmt.Sprintf("a write of %s from user id %d was refused: %s", req.Path, uid, a.Refused))
case a.Failed != "":
say(fmt.Sprintf("a write of %s from user id %d failed: %s", req.Path, uid, a.Failed))
default:
say(fmt.Sprintf("wrote %s for the provisioner's client (user id %d): %s, %s", req.Path, uid, a.Action,
a.Digest))
}
_ = conn.SetWriteDeadline(time.Now().Add(within))
reply(conn, a, say)
}
// busy answers the client's connection over atOnce: its request read and dropped first, briefly, so the client reads
// the answer rather than a broken pipe — while no more than busyWaiting such refusals wait at once; over that, at once.
func busy(conn *net.UnixConn, readBy time.Time, busyWaiting chan struct{}, say func(string)) {
select {
case busyWaiting <- struct{}{}:
by := time.Now().Add(refusedWithin)
if readBy.Before(by) {
by = readBy
}
_ = conn.SetReadDeadline(by)
drain(conn)
<-busyWaiting
default:
}
_ = conn.SetWriteDeadline(time.Now().Add(refusedWithin))
reply(conn, Refusal("", "busy: other writes are being served; ask again"), say)
}
// drain reads and drops a request up to its end — the newline after it, the end of the stream or the deadline —
// and at most MaxRequest bytes, keeping none of it.
func drain(r io.Reader) {
var buf [4096]byte
for read := 0; read < MaxRequest; {
n, err := r.Read(buf[:])
read += n
if bytes.IndexByte(buf[:n], '\n') >= 0 || err != nil {
return
}
}
}
func reply(conn *net.UnixConn, a Answer, say func(string)) {
defer conn.Close()
if err := json.NewEncoder(conn).Encode(a); err != nil {
say("a filled file's answer could not be written back: " + err.Error())
}
}
internal/filling/filling_test.go
-561
View File
@@ -1,561 +0,0 @@
package filling
import (
"context"
"encoding/json"
"errors"
"fmt"
"net"
"os"
"path/filepath"
"strings"
"sync"
"testing"
"time"
"github.com/novox/mesh-host/internal/declaration"
)
// Defends novox/hq ADR 0349 §4 (issue 668): the node-engine writes a filled file on request only for the provisioner
// client's account, only for a clean path the current declaration names as filled, and one at a time with the apply.
const provisionerUID = 4242
func declared(t *testing.T, raw string) *declaration.Declaration {
t.Helper()
d, err := declaration.Parse([]byte(raw))
if err != nil {
t.Fatal(err)
}
return d
}
const withAccount = `{"declaration":1,"provisioner_account":"mesh-node-provisioner","filled_files":["/etc/m/m.conf"],
"resources":[{"id":"m.conf","type":"file","path":"/etc/m/m.conf","content":"x","owner-at-runtime":"owned"},
{"id":"m.other","type":"file","path":"/etc/m/other.conf","content":"y"}]}`
const withoutAccount = `{"declaration":1,"filled_files":["/etc/m/m.conf"],
"resources":[{"id":"m.conf","type":"file","path":"/etc/m/m.conf","content":"x","owner-at-runtime":"owned"}]}`
// engine is an Engine over a fixed declaration, recording what it was asked to place.
type engine struct {
Engine
mu sync.Mutex
placed []string
}
func newEngine(t *testing.T, raw string, lock sync.Locker) *engine {
d := declared(t, raw)
e := &engine{}
e.Engine = Engine{
Declared: func() (*declaration.Declaration, error) { return d, nil },
UIDOf: func(name string) (uint32, error) {
if name == "mesh-node-provisioner" {
return provisionerUID, nil
}
return 0, errors.New("no such account")
},
Lock: func() (func(), error) { lock.Lock(); return lock.Unlock, nil },
Place: func(_ context.Context, _ *declaration.Declaration, path string, content []byte) (string, string, error) {
e.mu.Lock()
defer e.mu.Unlock()
e.placed = append(e.placed, path)
return "updated", "digest", nil
},
}
return e
}
func TestAWriteFromTheProvisionersAccountOfAFilledPathIsPlaced(t *testing.T) {
e := newEngine(t, withAccount, &sync.Mutex{})
if err := e.Accepts(provisionerUID); err != nil {
t.Fatal(err)
}
a := e.Write(context.Background(), provisionerUID, Request{Path: "/etc/m/m.conf", Content: []byte("x")})
if a.Refused != "" || a.Failed != "" || a.Action != "updated" || len(e.placed) != 1 {
t.Fatalf("answer %+v, placed %v", a, e.placed)
}
}
func TestAnUndeclaredPathIsRefused(t *testing.T) {
e := newEngine(t, withAccount, &sync.Mutex{})
for _, path := range []string{"/etc/m/other.conf", "/etc/m/none.conf", "/etc/passwd"} {
a := e.Write(context.Background(), provisionerUID, Request{Path: path, Content: []byte("evil")})
if a.Refused == "" || !strings.Contains(a.Refused, "not a file this node's declaration names as filled") {
t.Errorf("%s: %+v", path, a)
}
}
if len(e.placed) != 0 {
t.Errorf("placed %v", e.placed)
}
}
func TestAnotherAccountIsRefused(t *testing.T) {
e := newEngine(t, withAccount, &sync.Mutex{})
for _, uid := range []uint32{0, 1000, provisionerUID + 1} {
if err := e.Accepts(uid); err == nil {
t.Errorf("user id %d was accepted", uid)
}
a := e.Write(context.Background(), uid, Request{Path: "/etc/m/m.conf", Content: []byte("evil")})
if a.Refused == "" || !strings.Contains(a.Refused, "is not the provisioner client's account") {
t.Errorf("user id %d: %+v", uid, a)
}
}
if len(e.placed) != 0 {
t.Errorf("placed %v", e.placed)
}
}
func TestEveryWriteIsRefusedWhileTheDeclarationNamesNoAccount(t *testing.T) {
e := newEngine(t, withoutAccount, &sync.Mutex{})
if err := e.Accepts(provisionerUID); err == nil || !strings.Contains(err.Error(), "names no provisioner") {
t.Errorf("accepted with no account named: %v", err)
}
a := e.Write(context.Background(), provisionerUID, Request{Path: "/etc/m/m.conf", Content: []byte("x")})
if a.Refused == "" || len(e.placed) != 0 {
t.Errorf("written with no account named: %+v", a)
}
}
func TestAnAccountThatIsRootHereIsRefused(t *testing.T) {
e := newEngine(t, withAccount, &sync.Mutex{})
e.UIDOf = func(string) (uint32, error) { return 0, nil }
if err := e.Accepts(0); err == nil || !strings.Contains(err.Error(), "root") {
t.Errorf("an account that is root here was accepted: %v", err)
}
}
func TestAnUncleanPathIsRefused(t *testing.T) {
e := newEngine(t, withAccount, &sync.Mutex{})
for _, path := range []string{"", "etc/m/m.conf", "m.conf", "/etc//m/m.conf", "/etc/m/../m/m.conf",
"/etc/m/./m.conf", "/etc/m/m.conf/"} {
a := e.Write(context.Background(), provisionerUID, Request{Path: path, Content: []byte("x")})
if a.Refused == "" {
t.Errorf("%q was not refused: %+v", path, a)
}
}
if len(e.placed) != 0 {
t.Errorf("placed %v", e.placed)
}
}
// A write waits for an apply holding the apply's lock, and writes only once it is let go.
func TestAWriteWaitsForTheApply(t *testing.T) {
var applying sync.Mutex
e := newEngine(t, withAccount, &applying)
applying.Lock()
done := make(chan Answer)
go func() {
done <- e.Write(context.Background(), provisionerUID, Request{Path: "/etc/m/m.conf", Content: []byte("x")})
}()
select {
case a := <-done:
t.Fatalf("written while the apply held the lock: %+v", a)
case <-time.After(100 * time.Millisecond):
}
e.mu.Lock()
n := len(e.placed)
e.mu.Unlock()
if n != 0 {
t.Fatalf("placed while the apply held the lock")
}
applying.Unlock()
if a := <-done; a.Action != "updated" {
t.Fatalf("after the apply: %+v", a)
}
}
// fixed is a Writer that accepts one user id and answers every write.
type fixed struct {
uid uint32
mu sync.Mutex
writes []Request
}
func (f *fixed) Accepts(uid uint32) error {
if uid != f.uid {
return errors.New("not the provisioner client's account")
}
return nil
}
func (f *fixed) Write(_ context.Context, _ uint32, req Request) Answer {
f.mu.Lock()
defer f.mu.Unlock()
f.writes = append(f.writes, req)
return Answer{Path: req.Path, Action: "updated", Digest: "d"}
}
// serve serves w on a socket in a fresh directory and returns its path, and what was said.
func serve(t *testing.T, w Writer) (string, func() []string) {
t.Helper()
dir, err := os.MkdirTemp("", "fill")
if err != nil {
t.Fatal(err)
}
t.Cleanup(func() { os.RemoveAll(dir) })
socket := filepath.Join(dir, "fill.sock")
ctx, cancel := context.WithCancel(context.Background())
var mu sync.Mutex
var said []string
stopped := make(chan error)
go func() {
stopped <- Serve(ctx, socket, w, func(s string) { mu.Lock(); said = append(said, s); mu.Unlock() })
}()
t.Cleanup(func() {
cancel()
if err := <-stopped; err != nil {
t.Error(err)
}
})
for i := 0; i < 100; i++ {
if _, err := os.Stat(socket); err == nil {
break
}
time.Sleep(10 * time.Millisecond)
}
return socket, func() []string { mu.Lock(); defer mu.Unlock(); return append([]string(nil), said...) }
}
func ask(t *testing.T, socket string, req Request) Answer {
t.Helper()
conn, err := net.Dial("unix", socket)
if err != nil {
t.Fatal(err)
}
defer conn.Close()
// Bounded: a socket that never answers fails the test instead of hanging it.
_ = conn.SetDeadline(time.Now().Add(5 * time.Second))
if err := json.NewEncoder(conn).Encode(req); err != nil {
t.Fatal(err)
}
var a Answer
if err := json.NewDecoder(conn).Decode(&a); err != nil {
t.Fatal(err)
}
return a
}
// Over a real socket, the kernel says who is asking: this test's own account is the provisioner's here.
func TestTheSocketWritesForTheAccountTheKernelSays(t *testing.T) {
w := &fixed{uid: uint32(os.Getuid())}
socket, said := serve(t, w)
a := ask(t, socket, Request{Path: "/etc/m/m.conf", Content: []byte("secret value")})
if a.Refused != "" || a.Action != "updated" || len(w.writes) != 1 || string(w.writes[0].Content) != "secret value" {
t.Fatalf("answer %+v, writes %+v", a, w.writes)
}
for _, line := range said() {
if strings.Contains(line, "secret value") {
t.Errorf("the content was said: %q", line)
}
}
}
func TestTheSocketRefusesAnotherAccountBeforeReadingItsRequest(t *testing.T) {
w := &fixed{uid: uint32(os.Getuid()) + 1}
socket, said := serve(t, w)
conn, err := net.Dial("unix", socket)
if err != nil {
t.Fatal(err)
}
defer conn.Close()
// Closed at once, its request never read: the write may meet a closed socket, and the refusal is there to read.
_ = json.NewEncoder(conn).Encode(Request{Path: "/etc/m/m.conf", Content: []byte("secret value")})
var a Answer
if err := json.NewDecoder(conn).Decode(&a); err != nil {
t.Fatal(err)
}
if a.Refused == "" || len(w.writes) != 0 {
t.Fatalf("another account was written for: %+v", a)
}
for _, line := range said() {
if strings.Contains(line, "secret value") {
t.Errorf("the content was said: %q", line)
}
}
}
// floodersAreAnother has the kernel's word say another account for every connection a test binds to a name holding
// "flood": the test runs as one account, and the socket must tell the client from everyone else.
func floodersAreAnother(t *testing.T) {
t.Helper()
was := peerOf
peerOf = func(conn *net.UnixConn) (uint32, int32, error) {
uid, pid, err := was(conn)
if a, ok := conn.RemoteAddr().(*net.UnixAddr); ok && a != nil && strings.Contains(a.Name, "flood") {
uid++
}
return uid, pid, err
}
t.Cleanup(func() { peerOf = was })
}
// A flood of connections from another account, each saying nothing and held open, does not stop the client's write:
// each is refused and closed before it holds anything.
func TestAFloodFromAnotherAccountDoesNotStopTheClient(t *testing.T) {
floodersAreAnother(t)
w := &fixed{uid: uint32(os.Getuid())}
socket, _ := serve(t, w)
from := t.TempDir()
var flood []*net.UnixConn
t.Cleanup(func() {
for _, c := range flood {
c.Close()
}
})
for i := 0; i < 3*atOnce+busyWaitingCap(); i++ {
c, err := net.DialUnix("unix", &net.UnixAddr{Name: filepath.Join(from, fmt.Sprintf("flood-%d", i)), Net: "unix"},
&net.UnixAddr{Name: socket, Net: "unix"})
if err != nil {
t.Fatal(err)
}
flood = append(flood, c)
}
start := time.Now()
a := ask(t, socket, Request{Path: "/etc/m/m.conf", Content: []byte("x")})
if a.Refused != "" || a.Action != "updated" {
t.Fatalf("the client's write during a flood: %+v", a)
}
if took := time.Since(start); took > refusedWithin/2 {
t.Errorf("the client's write waited %s behind the flood", took)
}
// And every flooder was closed: it reads its refusal and then the end.
for _, c := range flood {
_ = c.SetReadDeadline(time.Now().Add(2 * time.Second))
var got Answer
if err := json.NewDecoder(c).Decode(&got); err != nil || got.Refused == "" {
t.Fatalf("a flooder was not refused at once: %+v %v", got, err)
}
}
}
func busyWaitingCap() int { return busyWaitingAtOnce }
// A client that connects and says nothing holds one slot only until within, and another write goes on meanwhile.
func TestASilentClientDoesNotHoldTheSocket(t *testing.T) {
was := within
within = 300 * time.Millisecond
t.Cleanup(func() { within = was })
w := &fixed{uid: uint32(os.Getuid())}
socket, _ := serve(t, w)
silent, err := net.Dial("unix", socket)
if err != nil {
t.Fatal(err)
}
defer silent.Close()
if a := ask(t, socket, Request{Path: "/etc/m/m.conf", Content: []byte("x")}); a.Action != "updated" {
t.Fatalf("a write beside a silent client: %+v", a)
}
_ = silent.SetReadDeadline(time.Now().Add(3 * time.Second))
var got Answer
if err := json.NewDecoder(silent).Decode(&got); err != nil || got.Refused == "" {
t.Errorf("the silent client was not let go after within: %+v %v", got, err)
}
}
// A request carrying more than the path and the content is refused: it says what the declaration says.
func TestARequestWithAnUnknownFieldIsRefused(t *testing.T) {
w := &fixed{uid: uint32(os.Getuid())}
socket, _ := serve(t, w)
conn, err := net.Dial("unix", socket)
if err != nil {
t.Fatal(err)
}
defer conn.Close()
if _, err := conn.Write([]byte(`{"path":"/etc/m/m.conf","content":"eA==","mode":"0777"}` + "\n")); err != nil {
t.Fatal(err)
}
var a Answer
if err := json.NewDecoder(conn).Decode(&a); err != nil {
t.Fatal(err)
}
if a.Refused == "" || !strings.Contains(a.Refused, "mode") || len(w.writes) != 0 {
t.Errorf("a request with an unknown field: %+v, writes %v", a, w.writes)
}
}
// An accept that fails is said, waited for and tried again: the socket goes on serving.
func TestTheSocketGoesOnAfterAnAcceptFails(t *testing.T) {
wasAccept, wasBackoff := acceptFrom, acceptBackoff
failures := 3
var mu sync.Mutex
acceptFrom = func(l *net.UnixListener) (*net.UnixConn, error) {
mu.Lock()
defer mu.Unlock()
if failures > 0 {
failures--
return nil, errors.New("accept: too many open files")
}
return l.AcceptUnix()
}
acceptBackoff = time.Millisecond
t.Cleanup(func() { acceptFrom, acceptBackoff = wasAccept, wasBackoff })
w := &fixed{uid: uint32(os.Getuid())}
socket, said := serve(t, w)
if a := ask(t, socket, Request{Path: "/etc/m/m.conf", Content: []byte("x")}); a.Action != "updated" {
t.Fatalf("after failed accepts: %+v", a)
}
if !strings.Contains(strings.Join(said(), "\n"), "too many open files") {
t.Errorf("the failed accepts were not said: %v", said())
}
}
// The socket's directory is given the mode the socket needs, as mesh-cli's is.
func TestTheSocketsDirectoryIsOpenedToEveryAccount(t *testing.T) {
dir, err := os.MkdirTemp("", "fill")
if err != nil {
t.Fatal(err)
}
t.Cleanup(func() { os.RemoveAll(dir) })
if err := os.Chmod(dir, 0o700); err != nil {
t.Fatal(err)
}
ctx, cancel := context.WithCancel(context.Background())
stopped := make(chan error)
go func() { stopped <- Serve(ctx, filepath.Join(dir, "fill.sock"), &fixed{}, nil) }()
for i := 0; i < 100; i++ {
if _, err := os.Stat(filepath.Join(dir, "fill.sock")); err == nil {
break
}
time.Sleep(10 * time.Millisecond)
}
cancel()
if err := <-stopped; err != nil {
t.Fatal(err)
}
info, err := os.Stat(dir)
if err != nil {
t.Fatal(err)
}
if info.Mode().Perm() != 0o755 {
t.Errorf("the socket's directory is %o", info.Mode().Perm())
}
}
// The first refusal of each kind for each account is said at once; what follows within the time is counted, and the
// count is said when the time is up, even when the account has stopped.
func TestRefusalsAreSaidOnceAndTheirCountIsNeverLost(t *testing.T) {
var mu sync.Mutex
var said []string
r := &refusalWords{say: func(s string) { mu.Lock(); said = append(said, s); mu.Unlock() }}
for i := 0; i < 3; i++ {
r.said(1000, "account", "not the provisioner client's account")
}
// Another kind from the same account is said at once, not hidden behind the first.
r.said(1000, "namespace", "another namespace")
if len(said) != 2 || !strings.Contains(said[0], "(account)") || !strings.Contains(said[1], "(namespace)") {
t.Fatalf("said at once: %q", said)
}
// The account stops; the time is up; the two counted are said.
r.flush()
if len(said) != 3 || !strings.Contains(said[2], "2 more") || !strings.Contains(said[2], "user id 1000") {
t.Fatalf("after the time: %q", said)
}
// And the next refusal is said at once again.
r.said(1000, "account", "not the provisioner client's account")
if len(said) != 4 {
t.Fatalf("after the count was said: %q", said)
}
}
// refusingSocket serves a socket that refuses this test's account, with counted refusals said every every. It returns
// how to be refused n times, everything said so far, and how to stop it.
func refusingSocket(t *testing.T, every time.Duration) (func(int), func() string, func()) {
t.Helper()
was := refusalsSaidEvery
refusalsSaidEvery = every
dir, err := os.MkdirTemp("", "fill")
if err != nil {
t.Fatal(err)
}
socket := filepath.Join(dir, "fill.sock")
ctx, cancel := context.WithCancel(context.Background())
var mu sync.Mutex
var said []string
all := func() string { mu.Lock(); defer mu.Unlock(); return strings.Join(said, "\n") }
stopped := make(chan error)
go func() {
stopped <- Serve(ctx, socket, &fixed{uid: uint32(os.Getuid()) + 1},
func(s string) { mu.Lock(); said = append(said, s); mu.Unlock() })
}()
for i := 0; i < 100; i++ {
if _, err := os.Stat(socket); err == nil {
break
}
time.Sleep(10 * time.Millisecond)
}
refuse := func(n int) {
for i := 0; i < n; i++ {
conn, err := net.Dial("unix", socket)
if err != nil {
t.Fatal(err)
}
_ = conn.SetDeadline(time.Now().Add(5 * time.Second))
var a Answer
_ = json.NewDecoder(conn).Decode(&a)
conn.Close()
}
}
var once sync.Once
stop := func() {
once.Do(func() {
cancel()
if err := <-stopped; err != nil {
t.Error(err)
}
refusalsSaidEvery = was
os.RemoveAll(dir)
})
}
t.Cleanup(stop)
return refuse, all, stop
}
// Refusals counted while the socket serves are said on its timer, though the account has stopped asking.
func TestTheSocketSaysItsCountedRefusalsOnATimer(t *testing.T) {
refuse, all, _ := refusingSocket(t, 200*time.Millisecond)
refuse(3)
// "more writes", not a number: a tick may fall between two refusals and split the count.
deadline := time.Now().Add(3 * time.Second)
for !strings.Contains(all(), "more writes") && time.Now().Before(deadline) {
time.Sleep(20 * time.Millisecond)
}
if !strings.Contains(all(), "more writes") {
t.Fatalf("the counted refusals were not said on the timer: %s", all())
}
}
// Refusals counted when the socket stops are said as it stops.
func TestTheSocketSaysItsCountedRefusalsWhenItStops(t *testing.T) {
refuse, all, stop := refusingSocket(t, time.Hour)
refuse(5)
if strings.Contains(all(), "4 more") {
t.Fatalf("said before the socket stopped: %s", all())
}
stop()
if !strings.Contains(all(), "4 more") {
t.Errorf("the refusals counted when the socket stopped were lost: %s", all())
}
}
// A connection over the bound of those waiting to be judged is closed unread, and said.
func TestAConnectionOverTheWaitingBoundIsClosedUnread(t *testing.T) {
was := judgeWaiting
judgeWaiting = 0
t.Cleanup(func() { judgeWaiting = was })
w := &fixed{uid: uint32(os.Getuid())}
socket, said := serve(t, w)
conn, err := net.Dial("unix", socket)
if err != nil {
t.Fatal(err)
}
defer conn.Close()
_ = conn.SetDeadline(time.Now().Add(5 * time.Second))
var a Answer
if err := json.NewDecoder(conn).Decode(&a); err == nil {
t.Fatalf("answered over the bound: %+v", a)
}
if len(w.writes) != 0 || !strings.Contains(strings.Join(said(), "\n"), "(overflow)") {
t.Errorf("writes %v, said %q", w.writes, said())
}
}
internal/identity/sealing.go
-64
View File
@@ -3,15 +3,11 @@ package identity
import (
"crypto/ecdh"
"crypto/rand"
"crypto/x509"
"encoding/base64"
"encoding/json"
"encoding/pem"
"fmt"
"os"
"strings"
"git.novox.be/novox/mesh-sdk/go/seal"
"golang.org/x/crypto/nacl/box"
)
@@ -145,63 +141,3 @@ func Seal(publicKey string, value []byte) (string, error) {
}
return base64.StdEncoding.EncodeToString(sealed), nil
}
// **The vault's box, opened with the same key** (novox/hq ADR 0349 §16, ADR 0206, issue 755). The core's own
// credentials travel in the declaration as values the vault sealed to this machine, which the controller carries and
// cannot open. The vault seals in the mesh's sealed box (mesh-sdk go/seal), whose keys travel as PEM: the public half
// as SPKI, the private as PKCS #8. They are this key's two halves in another wrapping, never another key.
// PublicPEM is this machine's sealing key's public half as SPKI PEM, the form the vault seals to.
func (s SealingKey) PublicPEM() (string, error) {
raw, err := base64.StdEncoding.DecodeString(s.Public)
if err != nil || len(raw) != 32 {
return "", fmt.Errorf("this node's sealing key is unusable")
}
pub, err := ecdh.X25519().NewPublicKey(raw)
if err != nil {
return "", fmt.Errorf("this node's sealing key is unusable: %w", err)
}
der, err := x509.MarshalPKIXPublicKey(pub)
if err != nil {
return "", err
}
return string(pem.EncodeToMemory(&pem.Block{Type: "PUBLIC KEY", Bytes: der})), nil
}
// OpenVaultBox opens a value the vault sealed to this machine. It fails for a box sealed to another key, or changed.
func (s SealingKey) OpenVaultBox(b seal.Box) ([]byte, error) {
raw, err := base64.StdEncoding.DecodeString(s.Private)
if err != nil || len(raw) != 32 {
return nil, fmt.Errorf("this node's sealing key is unusable")
}
priv, err := ecdh.X25519().NewPrivateKey(raw)
if err != nil {
return nil, fmt.Errorf("this node's sealing key is unusable: %w", err)
}
der, err := x509.MarshalPKCS8PrivateKey(priv)
if err != nil {
return nil, err
}
block := &pem.Block{Type: "PRIVATE KEY", Bytes: der}
opened, err := seal.Open(b, string(pem.EncodeToMemory(block)))
for i := range block.Bytes {
block.Bytes[i] = 0
}
if err != nil {
return nil, fmt.Errorf("this was not sealed by the vault to this node's current key")
}
return []byte(opened), nil
}
// Open opens anything the mesh sealed to this machine: the vault's box, given as its JSON (it begins with `{`), or
// the controller's sealed value, base64 (which never does).
func (s SealingKey) Open(sealed string) ([]byte, error) {
if strings.HasPrefix(strings.TrimSpace(sealed), "{") {
var b seal.Box
if err := json.Unmarshal([]byte(sealed), &b); err != nil {
return nil, fmt.Errorf("this is not the vault's sealed box: %w", err)
}
return s.OpenVaultBox(b)
}
return s.Unseal(sealed)
}
internal/link/handover.go
+1 -8
View File
@@ -143,14 +143,7 @@ func believed(check HandOverCheck, context string, withPath bool, body []byte) (
// believedBytes is believed with the ask's bytes as signed, for an ask that carries more than a hand-over's
// fields: read out of them only once they are believed.
func believedBytes(check HandOverCheck, context string, withPath bool, body []byte) (HandOverAsk, []byte, string) {
return believedBytesWithin(check, context, withPath, body, MaxHandOverAsk)
}
// believedBytesWithin is believedBytes for an ask bounded at max bytes: a move carries a sealed value, more than
// a hand-over's few hundred bytes (movesecret.go).
func believedBytesWithin(check HandOverCheck, context string, withPath bool, body []byte, max int) (HandOverAsk,
[]byte, string) {
if len(body) > max {
if len(body) > MaxHandOverAsk {
return HandOverAsk{}, nil, fmt.Sprintf("the ask is %d bytes, more than an ask", len(body))
}
var signed SignedHandOver
internal/link/hearing_nats.go
-23
View File
@@ -68,8 +68,6 @@ type natsLink struct {
judging *nats.Subscription
// acting hears the controller asking for the take act on a library (novox/hq ADR 0328 §7).
acting *nats.Subscription
// moving hears the controller asking to move a value sealed to this machine into the vault (issue 676).
moving *nats.Subscription
node string
// signer is the mesh's key, which a hand-over ask is verified with as a declaration is (issue 356).
signer ed25519.PublicKey
@@ -281,24 +279,6 @@ func (l *natsLink) HearAccessActs(handle AccessActFunc) {
}
}
// HearMoveSecrets listens for the controller asking this node to move one value sealed to it into the vault (design
// 57's V3, issue 676), on this machine's own subject, verified as a hand-over is (AnswerMoveSecret). A bus whose
// user list is older than the grant refuses the subscription; the controller then hears no answer.
func (l *natsLink) HearMoveSecrets(handle MoveSecretFunc) {
if handle == nil {
return
}
check := HandOverCheck{Signer: l.signer, Node: l.node}
moving, err := l.conn.Subscribe(AskMoveSecretSubject(l.node), func(msg *nats.Msg) {
if body, answer := MoveSecretReply(check, handle, msg.Reply, msg.Data); answer {
_ = msg.Respond(body)
}
})
if err == nil {
l.moving = moving
}
}
func (l *natsLink) Declarations() <-chan Declaration { return l.arrived }
func (l *natsLink) AskedToReport() <-chan struct{} { return l.asked }
func (l *natsLink) Lost() <-chan error { return l.lost }
@@ -319,9 +299,6 @@ func (l *natsLink) Close() {
if l.acting != nil {
_ = l.acting.Unsubscribe()
}
if l.moving != nil {
_ = l.moving.Unsubscribe()
}
if l.judging != nil {
_ = l.judging.Unsubscribe()
}
internal/link/messages.go
-61
View File
@@ -174,14 +174,6 @@ type Report struct {
// broken.
Windows []Window `json:"windows,omitempty"`
// Busy is every container the apply that reports left running because it said busy, and every one it
// replaced after it had waited, with how (novox/hq ADR 0335): `waiting`, `replaced-when-idle` or
// `replaced-at-bound`, since when, until when at the latest, and the last output of its busy command.
// A walk waiting on this machine for a module whose container is listed `waiting` waits for the report
// that no longer lists it so: its first-machine gate starts then (ADR 0335 §4). Absent from an engine
// older than BusyContract, which never waits.
Busy []Busy `json:"busy,omitempty"`
// Strays is what runs on the machine that the mesh neither wrote nor holds (novox/hq ADR
// 0163): containers nobody declared and nobody holds, the ones a cutover leaves behind.
Strays []Stray `json:"strays,omitempty"`
@@ -288,46 +280,6 @@ const LibraryContract = 6
// before it, its presence is what tells the controller this host may be sent `runner` on an identity.
const OwnGroupContract = 7
// BusyContract is the statement of a host that also reads a container's `busy` (novox/hq ADR 0335): it asks a
// container that declares it before replacing it, leaves it running while it says busy, within a bound of
// 15 minutes, and says the wait in its report's `busy`. Like the fields before it, its presence is what tells
// the controller this host may be sent `busy` on a container; an older host refuses the whole declaration
// for it.
const BusyContract = 8
// OwnershipContract is the statement of a host that also reads a file's `owner-at-runtime` and a declaration's
// `filled_files` (novox/hq ADR 0349 §5): it writes an owned file back on drift, a seed once and a managed file's
// keys or block only, exactly as their old forms, and refuses a filled path that is not a declared file saying
// its owner. Like the fields before it, its presence is what tells the controller this host may be sent those
// fields; an older host refuses the whole declaration for them.
const OwnershipContract = 9
// ProvisionerContract is the statement of a host that also reads a declaration's `provisioner_account` (novox/hq ADR
// 0349 §4) and serves the local write socket: it accepts a write of a filled file from that account alone. Its
// presence is what tells the controller this host may be sent the field; an older host refuses the whole
// declaration for it.
const ProvisionerContract = 10
// HandedContract is the statement of a host that writes a filled file only on the provisioner client's word (novox/hq
// ADR 0349 §5, design 57 P7): an apply neither writes nor records a file the declaration names as filled, and removes
// one as an orphan only when the write verb placed it. Its presence is what tells the controller it may hand a moved
// module's files to the provisioner here (mesh-controller internal/link/protocol.go's HandedContract): to an older
// host a filled file would have two writers.
const HandedContract = 11
// VaultBoxContract is the statement of a host that also reads a file's `vault-sealed` (novox/hq ADR 0349 §16, issue
// 755): a core module's own credential the vault sealed to this machine in the mesh's sealed box (ADR 0206), which
// the controller carries and cannot open, opened here with this machine's sealing key and written as a sealed file
// is. Its presence is what tells the controller it may send the vault's box here (mesh-controller internal/link
// protocol.go's VaultBoxContract); an older host refuses the whole declaration for the field, and is sent the
// controller's sealed value as before.
const VaultBoxContract = 12
// NodeKeyContract is the statement of a host that states its own sealing key to the vault's node-key, in its own name,
// at its start and whenever the key changes (novox/hq ADR 0372, issue 786). The same number as mesh-controller
// internal/link/protocol.go's NodeKeyContract.
const NodeKeyContract = 13
// Health is one statement of every long-running resource's state on this machine (to-be 48 §4). Said in
// every report, as an event on each change, and again every minute while anything is not healthy — so a
// lost statement is not a lost fault.
@@ -392,8 +344,6 @@ type UnitsHealth struct {
Failed []FailedUnit `json:"failed"`
// Unread names a manager that could not be read, with why.
Unread []string `json:"unread,omitempty"`
// Away names an account's own manager not running, so not read: nothing in it is failed (issue 764).
Away []string `json:"away,omitempty"`
}
// FailedUnit is one failed unit no module places.
@@ -781,17 +731,6 @@ type Window struct {
Until time.Time `json:"until"`
}
// Busy is one container a busy command held (novox/hq ADR 0335), as a report says it. ID is the resource's
// id, `<module>.<resource>`.
type Busy struct {
ID string `json:"id"`
Container string `json:"container"`
State string `json:"state"`
Since time.Time `json:"since"`
Until time.Time `json:"until"`
Said string `json:"said,omitempty"`
}
// A Stray is a container the mesh neither wrote nor holds (ADR 0163).
type Stray struct {
Kind string `json:"kind"`
internal/link/movesecret.go
-209
View File
@@ -1,209 +0,0 @@
package link
import (
"context"
"encoding/json"
"errors"
"fmt"
"strings"
"time"
)
// The move of a secret into the vault, asked over the bus (design 57's V3, issue 676; the operator's decision of
// 2026-10-14 on that issue).
//
// **Only the machine a value was sealed to can open it**, so the controller, which cannot, asks each machine's
// engine to move what was sealed to it: one row of its store at a time, on AskMoveSecretSubject, signed as a
// hand-over is and held to the same checks (believed), under a signing context of its own, naming the row — a
// module's own secret on this machine, or a pair credential with this machine at one end — its origin, the
// controller's generation, when it was made, and the value as the store holds it, sealed to this machine. The
// engine opens it with its sealing key, hands it to the vault's `move-in` in the vault's two steps, and answers
// what the vault kept only once the vault's read-back opens to the same value. Nothing is rotated; the engine
// writes nothing on this machine. The controller holds the same shapes in its own link code, with a test on each
// side holding the field names.
// AskMoveSecretSubject is where the controller asks this node's engine to move one value sealed to it into the vault.
func AskMoveSecretSubject(node string) string { return "mesh.node." + node + ".ask.move-secret" }
// MoveSecretContext is prefixed to a move ask's bytes before signing: never another ask's, nor a declaration's.
const MoveSecretContext = "novox-mesh move-secret v1\n"
// MoveSecretAsk is a move's own fields, beside a hand-over's (node, by, expires, nonce; no path): the row — module
// and name for a module's own secret on the ask's node; provider, provision, consumer, consumer module and local
// name for a pair credential — where its value came from, the controller's generation, when it was made, the value
// sealed to this machine as the store holds it, and the sealing key the store says it was sealed to.
type MoveSecretAsk struct {
Module string `json:"module,omitempty"`
Name string `json:"name,omitempty"`
Provider string `json:"provider,omitempty"`
Provision string `json:"provision,omitempty"`
Consumer string `json:"consumer,omitempty"`
ConsumerModule string `json:"consumer_module,omitempty"`
Local string `json:"local,omitempty"`
Origin string `json:"origin"`
Generation string `json:"generation,omitempty"`
MadeAt time.Time `json:"made_at"`
Sealed string `json:"sealed"`
SealedTo string `json:"sealed_to,omitempty"`
}
// MoveSecretAnswer is the engine's answer: what the vault kept, or why nothing was moved — never both.
type MoveSecretAnswer struct {
Said string `json:"said,omitempty"`
Refused string `json:"refused,omitempty"`
Secret string `json:"secret,omitempty"`
Generation int `json:"generation,omitempty"`
Unchanged bool `json:"unchanged,omitempty"`
}
// MoveSecret is one move asked and believed, as the handler is given it: the row, this machine, and who asked.
type MoveSecret struct {
MoveSecretAsk
Node, By string
}
// Own says the row is a module's own secret rather than a pair credential.
func (m MoveSecret) Own() bool { return m.Module != "" || m.Name != "" }
// MovedSecret is what the vault kept: its name, its generation, and whether it held the value already.
type MovedSecret struct {
Secret string
Generation int
Unchanged bool
}
// MoveSecretFunc moves one value into the vault: what it kept, or why not.
type MoveSecretFunc func(MoveSecret) (MovedSecret, error)
// MaxMoveSecretAsk bounds a move's request: a row's names, a time, a value sealed and a signature.
const MaxMoveSecretAsk = 64 << 10
// movedOrigins are the origins the controller's store holds.
var movedOrigins = map[string]bool{"made": true, "accepted": true, "issued": true}
func refusedMove(why string) MoveSecretAnswer {
return MoveSecretAnswer{Refused: why + "; nothing was moved"}
}
// oneName is a short name of one line.
func oneName(s string) bool {
return s != "" && len(s) <= MaxActField && !strings.ContainsAny(s, "\r\n\x00/ ")
}
// AnswerMoveSecret is the engine's judgement of one move ask's bytes: believed as a hand-over is, then the row —
// one shape, this machine's, a value to open and an origin — and only then the handler.
func AnswerMoveSecret(check HandOverCheck, handle MoveSecretFunc, body []byte) MoveSecretAnswer {
if handle == nil {
return refusedMove("this engine moves no secret into the vault")
}
ask, signed, refused := believedBytesWithin(check, MoveSecretContext, false, body, MaxMoveSecretAsk)
if refused != "" {
return refusedMove(refused)
}
var move MoveSecretAsk
if err := json.Unmarshal(signed, &move); err != nil {
return refusedMove("the signed ask is not a move's ask: " + err.Error())
}
own := move.Module != "" || move.Name != ""
pair := move.Provider != "" || move.Provision != "" || move.Consumer != "" || move.ConsumerModule != "" ||
move.Local != ""
switch {
case own == pair:
return refusedMove("the ask names neither one module's own secret nor one pair credential")
case own && (!oneName(move.Module) || !oneName(move.Name)):
return refusedMove("the ask's module and name are not short names")
case pair && (!oneName(move.Provider) || !oneName(move.Provision) || !oneName(move.Consumer) ||
!oneName(move.ConsumerModule) || (move.Local != "" && !oneName(move.Local))):
return refusedMove("the ask's provider, provision, consumer, consuming module and local name are not short names")
case pair && move.Provider != ask.Node && move.Consumer != ask.Node:
return refusedMove(fmt.Sprintf("the pair is between %s and %s, and this is %s", move.Provider, move.Consumer,
ask.Node))
case !movedOrigins[move.Origin]:
return refusedMove(fmt.Sprintf("the origin %q is not made, accepted or issued", move.Origin))
case strings.TrimSpace(move.Sealed) == "":
return refusedMove("the ask carries no value sealed to this machine")
case len(move.Generation) > MaxActField || strings.ContainsAny(move.Generation, "\r\n\x00"):
return refusedMove("the ask's generation is not one short line")
}
moved, err := handle(MoveSecret{MoveSecretAsk: move, Node: ask.Node, By: ask.By})
if err != nil {
return refusedMove(err.Error())
}
said := fmt.Sprintf("%s is in the vault at generation %d, read back the same", moved.Secret, moved.Generation)
if moved.Unchanged {
said = fmt.Sprintf("the vault held %s already, the same value (generation %d)", moved.Secret, moved.Generation)
}
return MoveSecretAnswer{Said: said, Secret: moved.Secret, Generation: moved.Generation, Unchanged: moved.Unchanged}
}
// MoveSecretReply is what a move request is answered with, and whether to answer, as HandOverReply.
func MoveSecretReply(check HandOverCheck, handle MoveSecretFunc, reply string, body []byte) ([]byte, bool) {
if reply == "" {
return nil, false
}
answer, err := json.Marshal(AnswerMoveSecret(check, handle, body))
if err != nil {
return nil, false
}
return answer, true
}
// MoveSecrets is a link that hears the controller asking this node to move a value into the vault. Optional.
type MoveSecrets interface {
HearMoveSecrets(MoveSecretFunc)
}
// ---- asking the vault ------------------------------------------------------------------------------------------
// VaultSeat is the vault's seat; its verbs are asked on `mesh.seat.<seat>.call.<verb>.<caller>`, the subject that
// names its caller (novox/hq issue 365), which the bus lets only that user publish.
const VaultSeat = "mesh-vault"
// NodeEngineUser is the bus user of a machine's node-engine, as the controller's broker names it.
func NodeEngineUser(node string) string { return "node." + node }
// CalledVaultSubject is where this node's engine asks one of the vault's verbs in its own name: each dot of its
// user's name written `~`, as the controller's broker writes a caller's token.
func CalledVaultSubject(verb, node string) string {
return "mesh.seat." + VaultSeat + ".call." + verb + "." + strings.ReplaceAll(NodeEngineUser(node), ".", "~")
}
// VaultWithin is how long the engine gives the vault to answer one step.
const VaultWithin = 15 * time.Second
// AskVault asks one of the vault's verbs in this node's engine's own name over the link open now, and answers its
// result, or its refusal as an error.
func (q *Queue) AskVault(ctx context.Context, verb string, args any) (json.RawMessage, error) {
q.init()
q.mu.Lock()
bus := q.bus
q.mu.Unlock()
if bus == nil {
return nil, fmt.Errorf("%w: this machine has no link to the bus open now", ErrNotAsked)
}
asker, ok := bus.(ToolBus)
if !ok {
return nil, fmt.Errorf("%w: the link to the bus open now cannot ask the vault", ErrNotAsked)
}
body, err := json.Marshal(args)
if err != nil {
return nil, err
}
asking, cancel := context.WithTimeout(ctx, VaultWithin)
defer cancel()
reply, err := asker.Ask(asking, CalledVaultSubject(verb, q.Membership.Node), body)
if err != nil {
return nil, fmt.Errorf("the vault's %s could not be asked: %w", verb, err)
}
var env struct {
Result json.RawMessage `json:"result"`
Error string `json:"error"`
}
if err := json.Unmarshal(reply, &env); err != nil {
return nil, fmt.Errorf("the vault answered %s with something that is not a tool's answer", verb)
}
if env.Error != "" {
return nil, errors.New(env.Error)
}
return env.Result, nil
}
internal/link/movesecret_test.go
-141
View File
@@ -1,141 +0,0 @@
package link
import (
"crypto/ed25519"
"encoding/json"
"errors"
"fmt"
"strings"
"testing"
"time"
"github.com/nats-io/nats.go"
)
// The move's ask (design 57's V3, issue 676): a hand-over's fields and the row's own in one signed object, under a
// signing context of its own. The controller's side holds the same names and words.
func TestTheMovesAskKeepsItsFieldNamesAndContext(t *testing.T) {
body, _ := json.Marshal(MoveSecretAsk{Module: "m", Name: "n", Provider: "p", Provision: "q", Consumer: "c",
ConsumerModule: "cm", Local: "l", Origin: "made", Generation: "g", MadeAt: time.Unix(1, 0).UTC(),
Sealed: "s", SealedTo: "k"})
if got := keysOf(t, body); got != "consumer consumer_module generation local made_at module name origin provider "+
"provision sealed sealed_to" {
t.Fatalf("the move's fields are %q", got)
}
if AskMoveSecretSubject("ace") != "mesh.node.ace.ask.move-secret" {
t.Fatalf("the subject is %q", AskMoveSecretSubject("ace"))
}
if MoveSecretContext != "novox-mesh move-secret v1\n" {
t.Fatalf("the signing context is %q", MoveSecretContext)
}
answer, _ := json.Marshal(MoveSecretAnswer{Said: "s", Refused: "r", Secret: "own/a/b/c", Generation: 1, Unchanged: true})
if got := keysOf(t, answer); got != "generation refused said secret unchanged" {
t.Fatalf("the answer's fields are %q", got)
}
}
// signedMove is a move's ask as the controller makes it.
func signedMove(t *testing.T, key ed25519.PrivateKey, ask HandOverAsk, move MoveSecretAsk) []byte {
t.Helper()
raw, _ := json.Marshal(struct {
HandOverAsk
MoveSecretAsk
}{ask, move})
body, _ := json.Marshal(SignedHandOver{Ask: raw, Signature: ed25519.Sign(key, append([]byte(MoveSecretContext), raw...))})
return body
}
// A move is believed as every signed ask is, then names one row this machine holds — its own secret, or a pair
// credential with this machine at one end — with a value sealed to it and an origin the store has; only then is the
// handler given it. A hand-over's signature is not a move's.
func TestAMoveIsTakenOnlyForARowThisMachineHolds(t *testing.T) {
m, key := aMember(t)
now := time.Now()
var given []MoveSecret
handle := func(s MoveSecret) (MovedSecret, error) {
given = append(given, s)
return MovedSecret{Secret: "own/" + m.Node + "/postgres/superuser-password", Generation: 1}, nil
}
check := HandOverCheck{Signer: m.Signer, Node: m.Node, Nonces: &Nonces{}}
ask := func(nonce string) HandOverAsk {
return HandOverAsk{Node: m.Node, By: "the move", Expires: now.Add(HandOverGood), Nonce: nonce}
}
own := MoveSecretAsk{Module: "postgres", Name: "superuser-password", Origin: "made", Sealed: "c2VhbGVk"}
for why, body := range map[string][]byte{
"a hand-over's": signedAsk(t, key, ask(aNonce+"01")),
"no value": signedMove(t, key, ask(aNonce+"02"), MoveSecretAsk{Module: "postgres", Name: "x", Origin: "made"}),
"no origin": signedMove(t, key, ask(aNonce+"03"), MoveSecretAsk{Module: "postgres", Name: "x", Sealed: "c2Vh"}),
"both shapes": signedMove(t, key, ask(aNonce+"04"), MoveSecretAsk{Module: "postgres", Name: "x", Provider: m.Node,
Provision: "database", Consumer: "shanks", ConsumerModule: "gitea", Origin: "made", Sealed: "c2Vh"}),
"neither end": signedMove(t, key, ask(aNonce+"05"), MoveSecretAsk{Provider: "anchor", Provision: "database",
Consumer: "shanks", ConsumerModule: "gitea", Origin: "made", Sealed: "c2Vh"}),
"a path": signedMove(t, key, HandOverAsk{Node: m.Node, Path: "/etc", By: "x", Expires: now.Add(HandOverGood),
Nonce: aNonce + "06"}, own),
"two lines": signedMove(t, key, ask(aNonce+"07"), MoveSecretAsk{Module: "postgres", Name: "x\ny", Origin: "made",
Sealed: "c2Vh"}),
} {
if a := AnswerMoveSecret(check, handle, body); a.Refused == "" || !strings.HasSuffix(a.Refused, "nothing was moved") {
t.Errorf("%s: not refused, or refused without saying nothing was moved: %+v", why, a)
}
}
if len(given) != 0 {
t.Fatalf("a refused ask reached the move: %+v", given)
}
a := AnswerMoveSecret(check, handle, signedMove(t, key, ask(aNonce+"08"), own))
if a.Refused != "" || a.Secret != "own/"+m.Node+"/postgres/superuser-password" || a.Generation != 1 || a.Said == "" ||
len(given) != 1 || given[0].Name != "superuser-password" || given[0].By != "the move" {
t.Fatalf("a good ask: %+v %+v", a, given)
}
pair := MoveSecretAsk{Provider: "anchor", Provision: "database", Consumer: m.Node, ConsumerModule: "gitea",
Origin: "accepted", Sealed: "c2Vh"}
if a := AnswerMoveSecret(check, handle, signedMove(t, key, ask(aNonce+"09"), pair)); a.Refused != "" {
t.Fatalf("a pair this machine consumes was refused: %+v", a)
}
// What the move refuses is said, and nothing else.
refusing := func(MoveSecret) (MovedSecret, error) { return MovedSecret{}, errors.New("the vault said no") }
if a := AnswerMoveSecret(check, refusing, signedMove(t, key, ask(aNonce+"10"), own)); !strings.Contains(a.Refused,
"the vault said no") || a.Secret != "" {
t.Fatalf("a refused move: %+v", a)
}
}
// **Against a real server**: the move's signed ask on the node's own subject is answered with what the handler
// moved, and the engine asks the vault on the subject that names it, `node~<machine>` last.
func TestNatsAMoveIsAskedOfTheNodeAndTheVaultIsAskedInItsName(t *testing.T) {
conn, js := aBus(t)
m, key := aMember(t)
vault, err := conn.Subscribe("mesh.seat.mesh-vault.call.move-in.node~"+m.Node, func(msg *nats.Msg) {
_ = msg.Respond([]byte(`{"result":{"call":"c1","seal_to":"k"}}`))
})
if err != nil {
t.Fatal(err)
}
defer vault.Unsubscribe()
l := &natsLink{conn: conn, js: js, node: m.Node, signer: m.Signer}
q := &Queue{Membership: m}
q.attach(t.Context(), l)
l.HearMoveSecrets(func(s MoveSecret) (MovedSecret, error) {
raw, err := q.AskVault(t.Context(), "move-in", map[string]string{"module": s.Module})
if err != nil || !strings.Contains(string(raw), `"call":"c1"`) {
return MovedSecret{}, fmt.Errorf("the vault answered %s, %v", raw, err)
}
return MovedSecret{Secret: "own/" + s.Node + "/" + s.Module + "/" + s.Name, Generation: 1}, nil
})
defer l.Close()
if err := conn.Flush(); err != nil {
t.Fatal(err)
}
ask := HandOverAsk{Node: m.Node, By: "the move", Expires: time.Now().Add(HandOverGood), Nonce: strings.Repeat("ab", 16)}
reply, err := conn.Request(AskMoveSecretSubject(m.Node), signedMove(t, key, ask,
MoveSecretAsk{Module: "postgres", Name: "superuser-password", Origin: "made", Sealed: "c2Vh"}), 5*time.Second)
if err != nil {
t.Fatal(err)
}
var said MoveSecretAnswer
if err := json.Unmarshal(reply.Data, &said); err != nil || said.Secret != "own/"+m.Node+"/postgres/superuser-password" {
t.Fatalf("answered %s, %v", reply.Data, err)
}
if _, err := (&Queue{Membership: Membership{Node: "nowhere"}}).AskVault(t.Context(), "move-in", nil); !errors.Is(err, ErrNotAsked) {
t.Fatalf("a queue with no link asked the vault: %v", err)
}
}
internal/link/queue.go
+2 -5
View File
@@ -65,11 +65,8 @@ type Queue struct {
// AccessAct judges and runs the take act on an access's folder, asked by the controller on an approved
// warrant (novox/hq ADR 0328 §7). Nil answers every such ask with a refusal.
AccessAct AccessActFunc
// MoveSecret moves one value sealed to this machine into the vault, asked by the controller during the move of
// the secrets (design 57's V3, issue 676). Nil answers every such ask with a refusal.
MoveSecret MoveSecretFunc
Say Announce
Timeout time.Duration
Say Announce
Timeout time.Duration
once sync.Once
wake chan struct{}
internal/link/run.go
-4
View File
@@ -256,10 +256,6 @@ func serve(ctx context.Context, link Link, m Membership, queue *Queue, say Annou
if s, ok := link.(AccessActs); ok {
s.HearAccessActs(queue.AccessAct)
}
// And asking to move a value sealed to this machine into the vault (design 57's V3, issue 676).
if s, ok := link.(MoveSecrets); ok {
s.HearMoveSecrets(queue.MoveSecret)
}
for {
// Asked to stop — by standing aside, say — nothing further is taken in.
internal/liveness/liveness_test.go
-53
View File
@@ -330,56 +330,3 @@ func TestOneLookIsOneReadOfEveryContainer(t *testing.T) {
t.Fatalf("a container the inspect found, beside one it did not: %+v", s)
}
}
// **A look starts no account's manager** (novox/hq issue 764, ADR 0177): a user-scoped unit's manager is
// asked of the machine's first; one not running is not read, nothing is started, and its unit is said
// unknown with why — away, never down; one running is read as before.
func TestAUserUnitsManagerNotRunningIsNotReadAndNotDown(t *testing.T) {
var asked []string
manager := "inactive"
run := func(_ context.Context, name string, args ...string) (string, error) {
asked = append(asked, name+" "+strings.Join(args, " "))
switch {
case name == "getent":
return "operator:x:1000:1000::/home/operator:/bin/zsh\n", nil
case name == "systemctl" && len(args) > 0 && args[0] == "is-active":
if manager != "active" {
return manager + "\n", errors.New("systemctl exited 3")
}
return "active\n", nil
case name == "systemctl":
return "Id=dunst.service\nLoadState=loaded\nActiveState=active\nSubState=running\nNRestarts=0\nInvocationID=i1\n", nil
}
return "", errors.New("not a command the judge may run")
}
c := &clock{now: t0}
j := aJudge(t, &Exec{Run: run, System: "arch"}, c, "")
j.Set([]Resource{{Module: "dunst", ID: "dunst.daemon", Kind: KindService, Target: "dunst.service",
Scope: "user", User: "operator"}})
var st Statement
for i := 0; i < 4; i++ {
c.now = t0.Add(time.Duration(i) * time.Minute)
st, _ = j.Look(t.Context())
}
if s := stateOf(t, st, "dunst.daemon"); s.State != Unknown || !strings.Contains(s.Reason, "not logged in") {
t.Fatalf("a unit in a manager not running is away, not down: %+v", s)
}
for _, a := range asked {
if strings.Contains(a, "--machine=") {
t.Fatalf("the account's own manager was asked while it was not running: %q", a)
}
if a != "getent passwd operator" && a != "systemctl is-active user@1000.service" {
t.Errorf("asked something but whether the manager runs: %q", a)
}
}
// Logged in: read in the account's own manager, as before.
manager, asked = "active", nil
c.now = t0.Add(5 * time.Minute)
st, _ = j.Look(t.Context())
if s := stateOf(t, st, "dunst.daemon"); s.State == Unknown || s.State == Unhealthy {
t.Fatalf("a unit in a running manager is read: %+v", s)
}
if last := asked[len(asked)-1]; !strings.HasPrefix(last, "systemctl --user --machine=operator@ show ") {
t.Fatalf("the running manager is read in the account's own: %q", last)
}
}
internal/liveness/runtime.go
+1 -20
View File
@@ -8,8 +8,6 @@ import (
"strconv"
"strings"
"sync"
"github.com/novox/mesh-host/internal/system"
)
// Runner runs a command and answers what it printed — the apply's own (apply.ExecRunner), so a look
@@ -18,13 +16,9 @@ type Runner func(ctx context.Context, name string, args ...string) (string, erro
// Exec reads the container runtime and the service manager through their command lines: one inspect of
// every container, one show of every unit per manager. **Reads only** — the whole of what this file may
// ask either of them is `container inspect`, `version` and `show`, and of the machine's manager whether an
// account's own is active (ADR 0240 rule 6, and a test holds it).
// ask either of them is `container inspect`, `version` and `show` (ADR 0240 rule 6, and a test holds it).
type Exec struct {
Run Runner
// System is what the host was built for: it says how the machine's manager is asked whether an
// account's own runs (system.ManagerRunning).
System string
mu sync.Mutex
// cri is the runtime that answered, once one has.
@@ -152,19 +146,6 @@ func (e *Exec) Units(ctx context.Context, scope, user string, units []string) (m
}
args := []string{"show", "--property=" + unitProperties, "--"}
if scope == "user" && user != "" {
// **A look starts no account's manager** (novox/hq issue 764, ADR 0177): reading it is a login,
// which starts it when none runs. The machine's manager is asked first; a manager not running is
// away — its units unread, said unknown with why, never down — as the apply says them waiting.
running, exists, err := system.ManagerRunning(ctx, e.System, system.Runner(e.Run), user)
switch {
case err != nil:
return nil, fmt.Errorf("whether %s's own service manager runs could not be read: %w", user, err)
case !exists:
return nil, fmt.Errorf("the machine has no account %q, whose own manager these units are in", user)
case !running:
return nil, fmt.Errorf("%s's own service manager is not running: the account is not logged in "+
"and does not linger", user)
}
args = append([]string{"--user", "--machine=" + user + "@"}, args...)
}
said, err := e.Run(ctx, "systemctl", append(args, units...)...)
internal/meshcli/peer.go
-19
View File
@@ -1,19 +0,0 @@
package meshcli
import "net"
// The kernel's word on who is at the other end of a socket, shared with the node-engine's other local socket, the
// write of a filled file (internal/filling, novox/hq ADR 0349 §4): one reading of it, so the two never judge a peer
// differently.
// PeerCredentials is the user id and process id of the process at the other end, as the kernel recorded them when
// it connected (SO_PEERCRED). Nothing the process says is read.
func PeerCredentials(conn *net.UnixConn) (uint32, int32, error) { return peerCredentials(conn) }
// SameNamespaces refuses a process whose mount or user namespace is not the engine's: a user id is this machine's
// account only in this machine's namespaces.
func SameNamespaces(pid int32) error { return namespacesOf(pid) }
// NotOwnDirectory says what is wrong with a socket's directory, or nothing: it must be a directory, not a link,
// owned by the account the engine runs as.
func NotOwnDirectory(dir string) string { return notOwnDirectory(dir) }
internal/outward/networks.go
-92
View File
@@ -1,92 +0,0 @@
package outward
import (
"errors"
"fmt"
"net"
"net/netip"
"sort"
)
// Network is one network this machine is directly attached to, and the link it is on (novox/hq ADR
// 0340 §7): the address and prefix of the network, not of the machine's own address on it.
//
// **Every one is said, and the machine judges none of them.** Which count as the machine's local
// network — private ranges only, never a public subnet, never a link-local one — is the controller's
// judgement (ADR 0340 §2), made where the filter is composed and shown with why; a machine that left a
// network out would make that judgement silently.
type Network struct {
Link string `json:"link"`
Network string `json:"network"`
}
// AddressesOf is the addresses on one link, as address and prefix. A parameter so a test can hold a
// machine's links without one.
type AddressesOf func(link string) ([]netip.Prefix, error)
// RealAddresses reads one link's addresses from the kernel.
func RealAddresses(link string) ([]netip.Prefix, error) {
iface, err := net.InterfaceByName(link)
if err != nil {
return nil, err
}
addrs, err := iface.Addrs()
if err != nil {
return nil, err
}
var out []netip.Prefix
for _, a := range addrs {
ipnet, ok := a.(*net.IPNet)
if !ok {
continue
}
addr, ok := netip.AddrFromSlice(ipnet.IP)
if !ok {
continue
}
bits, _ := ipnet.Mask.Size()
out = append(out, netip.PrefixFrom(addr.Unmap(), bits))
}
return out, nil
}
// Networks is the networks on each of these links — the links that face outside (Links) — sorted by link
// and network, without repeats. A link that has gone since Links read it has no networks rather than
// failing the rest; a link whose addresses cannot be read fails, so the machine says the fact is unknown
// rather than reporting a partial account as the whole.
func Networks(links []string, addressesOf AddressesOf) ([]Network, error) {
if addressesOf == nil {
addressesOf = RealAddresses
}
seen := map[Network]bool{}
var out []Network
for _, link := range links {
prefixes, err := addressesOf(link)
if err != nil {
if isNoSuchInterface(err) {
continue
}
return nil, fmt.Errorf("cannot read the addresses on %s: %w", link, err)
}
for _, p := range prefixes {
n := Network{Link: link, Network: p.Masked().String()}
if !seen[n] {
seen[n] = true
out = append(out, n)
}
}
}
sort.Slice(out, func(i, j int) bool {
if out[i].Link != out[j].Link {
return out[i].Link < out[j].Link
}
return out[i].Network < out[j].Network
})
return out, nil
}
// isNoSuchInterface is the error net.InterfaceByName gives for a link that is not there.
func isNoSuchInterface(err error) bool {
var op *net.OpError
return errors.As(err, &op) && op.Err != nil && op.Err.Error() == "no such network interface"
}
internal/outward/networks_test.go
-59
View File
@@ -1,59 +0,0 @@
package outward
import (
"errors"
"net/netip"
"reflect"
"testing"
)
func prefixes(s ...string) []netip.Prefix {
var out []netip.Prefix
for _, p := range s {
out = append(out, netip.MustParsePrefix(p))
}
return out
}
// **Each link facing outside says the networks it is attached to** (novox/hq ADR 0340 §7), as the
// network's address and prefix, never the machine's own address on it — and every one, public and
// link-local included, because which count is the controller's judgement and said with why.
func TestEachOutwardLinkSaysTheNetworksItIsOn(t *testing.T) {
addrs := map[string][]netip.Prefix{
"eth0": prefixes("192.168.1.10/24", "fe80::1/64", "fd12:3456:789a::5/64"),
"eth1": prefixes("203.0.113.7/24"),
// Not asked: the bridge is not a link facing outside.
"docker0": prefixes("172.17.0.1/16"),
}
got, err := Networks([]string{"eth1", "eth0"}, func(link string) ([]netip.Prefix, error) { return addrs[link], nil })
if err != nil {
t.Fatal(err)
}
want := []Network{
{Link: "eth0", Network: "192.168.1.0/24"},
{Link: "eth0", Network: "fd12:3456:789a::/64"},
{Link: "eth0", Network: "fe80::/64"},
{Link: "eth1", Network: "203.0.113.0/24"},
}
if !reflect.DeepEqual(got, want) {
t.Fatalf("got %+v\nwant %+v", got, want)
}
}
// A link whose addresses cannot be read fails the whole, so the machine says the fact is unknown rather
// than reporting part of its networks as all of them.
func TestAnUnreadableLinkIsNotAPartialAccount(t *testing.T) {
_, err := Networks([]string{"eth0"}, func(string) ([]netip.Prefix, error) { return nil, errors.New("denied") })
if err == nil {
t.Fatal("an unreadable link was reported as having no networks")
}
}
// A link gone between reading the routing table and reading its addresses has no networks, and the
// rest are still said.
func TestALinkGoneSinceItWasListedHasNoNetworks(t *testing.T) {
got, err := Networks([]string{"gone0"}, RealAddresses)
if err != nil || len(got) != 0 {
t.Fatalf("got %v, %v", got, err)
}
}
internal/profile/machine.go
+2 -35
View File
@@ -10,8 +10,6 @@ import (
"sync"
"syscall"
"time"
"github.com/novox/mesh-host/internal/outward"
)
// Machine is what this machine is made of, as it says itself (novox/hq issue 160): what somebody
@@ -52,11 +50,6 @@ type Machine struct {
// Filesystems is each local filesystem mounted, by where, with its size and what is free.
Filesystems []Filesystem `json:"filesystems,omitempty"`
// Networks is every network the machine is directly attached to on a link facing outside it, as the
// link and the network's address and prefix (novox/hq ADR 0340 §7): what a port reached from
// `local-network` is resolved from. Every one is said; which count is the controller's judgement.
Networks []outward.Network `json:"networks,omitempty"`
// Unknown names each fact that could not be determined, and why.
Unknown map[string]string `json:"unknown,omitempty"`
}
@@ -86,12 +79,11 @@ const (
FactTimezone = "timezone"
FactFilesystems = "filesystems"
FactPublicAddress = "public address"
FactNetworks = "networks"
)
// MachineFacts is every fact Unknown may name, in the order a reader shows them.
var MachineFacts = []string{FactKernelRelease, FactDistribution, FactVersion, FactProcessor, FactCores,
FactThreads, FactMemory, FactVirtual, FactUptime, FactTimezone, FactFilesystems, FactNetworks, FactPublicAddress}
FactThreads, FactMemory, FactVirtual, FactUptime, FactTimezone, FactFilesystems, FactPublicAddress}
// publicAddressUnknown is why the public uplink address is not said. A machine behind a NAT cannot see
// its public address from itself, and asking a service outside the mesh is a dependency the mesh does not
@@ -109,24 +101,11 @@ type Sources struct {
// Statfs is (size, free) in bytes of the filesystem mounted at a path.
Statfs func(path string) (size, free int64, err error)
Now func() time.Time
// Networks is the networks on the links facing outside this machine (novox/hq ADR 0340).
Networks func() ([]outward.Network, error)
}
// RealSources reads this machine.
func RealSources() Sources {
return Sources{Read: os.ReadFile, Readlink: os.Readlink, Run: ExecRunner, Statfs: statfs, Now: time.Now,
Networks: attachedNetworks}
}
// attachedNetworks is this machine's networks on the links that face outside it, read the way the report
// reads those links (novox/hq ADR 0140), so the two never disagree about which links those are.
func attachedNetworks() ([]outward.Network, error) {
links, err := outward.Links("", "")
if err != nil {
return nil, err
}
return outward.Networks(links, nil)
return Sources{Read: os.ReadFile, Readlink: os.Readlink, Run: ExecRunner, Statfs: statfs, Now: time.Now}
}
func statfs(path string) (int64, int64, error) {
@@ -214,18 +193,6 @@ func DescribeMachine(ctx context.Context, src Sources) Machine {
}
}
// The networks it is attached to, for the local network (novox/hq ADR 0340): every one on a link facing
// outside, or why none is said — never an empty list that reads as a machine on no network.
if src.Networks == nil {
unknown(FactNetworks, "this node-engine was given no way to read its links")
} else if networks, err := src.Networks(); err != nil {
unknown(FactNetworks, err.Error())
} else if len(networks) == 0 {
unknown(FactNetworks, "no link facing outside this machine carries an address")
} else {
m.Networks = networks
}
unknown(FactPublicAddress, publicAddressUnknown)
return m
}
internal/profile/machine_test.go
-29
View File
@@ -10,8 +10,6 @@ import (
"sync/atomic"
"testing"
"time"
"github.com/novox/mesh-host/internal/outward"
)
// The parsing layer, with the machine's files given (novox/hq ADR 0017 allows a fake here and only here);
@@ -42,9 +40,6 @@ func given(read map[string]string) Sources {
Run: func(context.Context, string, ...string) (string, error) { return "none\n", errors.New("exited 1") },
Statfs: func(p string) (int64, int64, error) { return 1000, 400, nil },
Now: func() time.Time { return time.Date(2026, 10, 12, 15, 0, 0, 0, time.UTC) },
Networks: func() ([]outward.Network, error) {
return []outward.Network{{Link: "eth0", Network: "192.168.1.0/24"}, {Link: "eth1", Network: "203.0.113.0/24"}}, nil
},
}
}
@@ -97,7 +92,6 @@ func TestMachine_aFactItCannotReadIsSaidUnknownNeverZero(t *testing.T) {
src := given(map[string]string{})
src.Readlink = func(string) (string, error) { return "", os.ErrNotExist }
src.Run = func(context.Context, string, ...string) (string, error) { return "", errors.New("not found") }
src.Networks = func() ([]outward.Network, error) { return nil, errors.New("cannot read the routing table") }
m := DescribeMachine(context.Background(), src)
raw, _ := json.Marshal(m)
@@ -189,26 +183,3 @@ func TestAgainstThisMachine_machineSaysWhatItIs(t *testing.T) {
raw, _ := json.MarshalIndent(m, "", " ")
t.Logf("%s", raw)
}
// **The networks it is attached to are among its facts** (novox/hq ADR 0340 §7), every one — the public
// one too, which the controller judges and never counts — and a machine with none on a link facing
// outside says why rather than reporting an empty list.
func TestMachine_saysTheNetworksItIsAttachedTo(t *testing.T) {
m := DescribeMachine(context.Background(), given(files))
raw, _ := json.Marshal(m)
var said struct {
Networks []map[string]string `json:"networks"`
}
_ = json.Unmarshal(raw, &said)
if len(said.Networks) != 2 || said.Networks[0]["link"] != "eth0" || said.Networks[0]["network"] != "192.168.1.0/24" ||
said.Networks[1]["network"] != "203.0.113.0/24" {
t.Fatalf("networks said as %s", raw)
}
src := given(files)
src.Networks = func() ([]outward.Network, error) { return nil, nil }
m = DescribeMachine(context.Background(), src)
if len(m.Networks) != 0 || !strings.Contains(m.Unknown[FactNetworks], "no link facing outside") {
t.Fatalf("a machine with no network said %v, unknown %q", m.Networks, m.Unknown[FactNetworks])
}
}
internal/store/store.go
-5
View File
@@ -155,11 +155,6 @@ type Applied struct {
// container's spec label; this is the same information kept per file, so that when the spec
// no longer matches the host can say WHICH file changed rather than only that something did.
Reads map[string]string `json:"reads,omitempty"`
// BusySince is, for a container left running because it said busy, when the first apply that would
// have replaced it ran (novox/hq ADR 0335 §3): kept here, beside the machine's state, so the bound runs
// from then across every pass and across a restart of the node-engine. Absent while nothing waits.
BusySince *time.Time `json:"busy_since,omitempty"`
}
// Into is what a file written into held before the mesh's keys, or before the mesh's block.
internal/system/system.go
-21
View File
@@ -304,27 +304,6 @@ func For(name string) (System, error) {
name, strings.Join(names, ", "))
}
// userManagers is a service manager that runs a manager per account (novox/hq ADR 0177).
type userManagers interface {
UserManagerRunning(ctx context.Context, run Runner, account string) (running, exists bool, err error)
}
// ManagerRunning is whether account's own manager runs on a machine built for name, asked of the
// machine's manager through run — never of the account's, which a question would start (see
// arch.UserManagerRunning). A look asks it before it reads anything in an account's manager (novox/hq
// issue 764). A system with no manager per account is an error: nothing can be read there.
func ManagerRunning(ctx context.Context, name string, run Runner, account string) (running, exists bool, err error) {
sys, err := For(name)
if err != nil {
return false, false, err
}
um, ok := sys.(userManagers)
if !ok {
return false, false, fmt.Errorf("this machine's service manager (%s) has no manager per account", sys.Name())
}
return um.UserManagerRunning(ctx, run, account)
}
// All is every system the host can be built for.
func All() []System {
return []System{arch{}, alpine{}, android{}}
internal/units/exec.go
+2 -12
View File
@@ -5,8 +5,6 @@ import (
"errors"
"fmt"
"strings"
"github.com/novox/mesh-host/internal/system"
)
// Runner runs a command and answers what it printed — the apply's own (apply.ExecRunner), so a look reads
@@ -14,9 +12,8 @@ import (
type Runner func(ctx context.Context, name string, args ...string) (string, error)
// Exec reads the service managers through systemctl and the package database through the system's own
// owner query. **Reads only**: `list-units`, `show`, the owner query, and — of the machine's manager and
// user database — whether an account's manager is active, are the whole of what it asks (ADR 0240 rule 6,
// and a test holds it).
// owner query. **Reads only**: `list-units`, `show` and the owner query are the whole of what it asks
// (ADR 0240 rule 6, and a test holds it).
type Exec struct {
Run Runner
// System is what the host was built for; it says how a file's package is asked. A system with no
@@ -106,10 +103,3 @@ func (e Exec) PackageOwning(ctx context.Context, path string) (string, bool, err
}
return "", false, nil
}
// UserManagerRunning asks the machine's manager whether an account's own runs, as the apply does before it
// touches a user-scoped unit (system.UserManagerRunning; novox/hq ADR 0177, issue 764). A system with no
// manager per account cannot say, and is an error: that manager is said unread.
func (e Exec) UserManagerRunning(ctx context.Context, account string) (running, exists bool, err error) {
return system.ManagerRunning(ctx, e.System, system.Runner(e.Run), account)
}
internal/units/units.go
+12 -48
View File
@@ -9,8 +9,8 @@
// own mount table, and a unit a removed package left behind, failed beside it, said by nobody.
//
// On every look the engine asks each service manager the mesh places units in — the machine's, and the
// account manager of every account the declaration places a user-scoped unit for — which units failed,
// and says each one's owner:
// account manager of every account the declaration names — which units failed, and says each one's
// owner:
//
// 1. a service or process the declaration states, by its unit and manager;
// 2. a file the declaration writes, when the unit's file is that file;
@@ -23,12 +23,6 @@
//
// **It reads; it never acts** (ADR 0240 rule 6): `list-units`, `show` and the package manager's owner
// query are the whole of what it asks. It neither resets nor restarts anything.
//
// **A look starts no account's manager** (novox/hq issue 764, ADR 0177). Reading an account's manager is
// a login, and a login starts the manager when none runs; so the machine's own manager is asked first
// whether user@<uid>.service is active, and the account's is read only then. A manager that is not
// running is away: no unit in it is failed, none is read, and the statement says it away — the apply's
// word for the same absence ("waiting" for its manager).
package units
import (
@@ -78,9 +72,7 @@ type Owned struct {
// Files is keyed by path; Packages by package name, only those declared present.
Files map[string]owner
Packages map[string]owner
// Accounts are the accounts whose own managers are read, sorted: each one the declaration places a
// user-scoped unit for, and no other (novox/hq issue 764). An account the declaration only creates
// has no unit of the mesh's in its manager to fail, and asking that manager starts it.
// Accounts are the accounts whose own managers are read, sorted.
Accounts []string
}
@@ -131,6 +123,8 @@ func OwnedBy(d *declaration.Declaration, held map[string]bool) Owned {
if !v.Absent {
o.Packages[v.Package] = own
}
case *declaration.User:
accounts[v.Name] = true
}
}
for a := range accounts {
@@ -175,10 +169,6 @@ type Reader interface {
Show(ctx context.Context, scope, user string, units []string) (map[string]Shown, error)
// PackageOwning is the package a file belongs to; false when none does or the machine cannot say.
PackageOwning(ctx context.Context, path string) (string, bool, error)
// UserManagerRunning is whether an account's own manager runs now, asked of the machine's manager
// and never of the account's, which a question would start (novox/hq issue 764); exists is false
// for an account the machine does not have.
UserManagerRunning(ctx context.Context, account string) (running, exists bool, err error)
}
// Failed is one failed unit as the statement says it.
@@ -208,9 +198,6 @@ type Statement struct {
Failed []Failed
// Unread names each manager that could not be read, with why.
Unread []string
// Away names each account manager not running, so not read: its account is not logged in and does
// not linger. Nothing in it is failed (novox/hq issue 764).
Away []string
}
// Owned answers the failures a module owns; Unowned those no module does.
@@ -303,39 +290,16 @@ func (j *Judge) Look(ctx context.Context) (Statement, bool) {
st := Statement{At: now, State: Unknown}
read := 0
failedNow := map[string]bool{}
keep := func(m manager) {
// What it held is neither cleared nor counted while it cannot be read.
for k := range j.seen {
if strings.HasPrefix(k, key(m.scope, m.user, "")) {
failedNow[k] = true
}
}
}
for _, m := range managers {
if m.scope == ScopeUser {
// Asked of the machine's manager first: reading the account's own would start it.
running, exists, err := j.reader.UserManagerRunning(ctx, m.user)
switch {
case err != nil:
st.Unread = append(st.Unread, fmt.Sprintf("%s could not be read: %s", m, firstLine(err.Error())))
keep(m)
continue
case !exists:
st.Unread = append(st.Unread, fmt.Sprintf("%s could not be read: the machine has no account %q",
m, m.user))
keep(m)
continue
case !running:
// Away: nothing in it runs, so nothing in it is failed; what it held is cleared.
st.Away = append(st.Away, m.user+"'s own service manager is not running: the account is not "+
"logged in and does not linger, so its units are not read")
continue
}
}
listed, err := j.reader.Failed(ctx, m.scope, m.user)
if err != nil {
st.Unread = append(st.Unread, fmt.Sprintf("%s could not be read: %s", m, firstLine(err.Error())))
keep(m)
// What it held is neither cleared nor counted while it cannot be read.
for k := range j.seen {
if strings.HasPrefix(k, key(m.scope, m.user, "")) {
failedNow[k] = true
}
}
continue
}
read++
@@ -389,7 +353,7 @@ func (j *Judge) Look(ctx context.Context) (Statement, bool) {
if read == 0 {
st.State = Unknown
}
word := st.State + "|away:" + strings.Join(st.Away, ",")
word := st.State
for _, f := range st.Failed {
word += "|" + f.Scope + "/" + f.Unit + "/" + f.Module
}
internal/units/units_test.go
-117
View File
@@ -21,11 +21,6 @@ type fakeReader struct {
owners map[string]string // by path
unread map[string]error
ownerQs int
// away is each account whose own manager is not running; noAccount each the machine lacks.
away map[string]bool
noAccount map[string]bool
// read is every manager whose failed units were asked, in order.
read []string
}
func mgr(scope, user string) string {
@@ -36,7 +31,6 @@ func mgr(scope, user string) string {
}
func (f *fakeReader) Failed(_ context.Context, scope, user string) ([]Listed, error) {
f.read = append(f.read, mgr(scope, user))
if err := f.unread[mgr(scope, user)]; err != nil {
return nil, err
}
@@ -51,13 +45,6 @@ func (f *fakeReader) Show(_ context.Context, _, _ string, units []string) (map[s
return out, nil
}
func (f *fakeReader) UserManagerRunning(_ context.Context, account string) (bool, bool, error) {
if f.noAccount[account] {
return false, false, nil
}
return !f.away[account], true, nil
}
func (f *fakeReader) PackageOwning(_ context.Context, path string) (string, bool, error) {
f.ownerQs++
p, ok := f.owners[path]
@@ -88,11 +75,6 @@ func declared() *declaration.Declaration {
&declaration.Package{ID: "openrazer.daemon", Type: declaration.TypePackage, Package: "openrazer-daemon"},
&declaration.Package{ID: "clipmenu.greenclip", Type: declaration.TypePackage, Package: "rofi-greenclip", Absent: true},
&declaration.User{ID: "zsh.account", Type: declaration.TypeUser, Name: "operator"},
// The operator's own manager is read because a module places a unit in it (dunst, on a desktop).
&declaration.Service{ID: "dunst.reload", Type: declaration.TypeService, Unit: "dunst.service",
Scope: declaration.ScopeUser, User: "operator"},
// A module account the declaration only creates: its manager is never asked (issue 764).
&declaration.User{ID: "telegram.account", Type: declaration.TypeUser, Name: "telegram"},
&declaration.Service{ID: "sshd.run", Type: declaration.TypeService, Unit: "sshd.service", State: "running"},
&declaration.Service{ID: "power.after-boot", Type: declaration.TypeService, Unit: "mesh-power-after-boot.service"},
&declaration.File{ID: "watcher.unit", Type: declaration.TypeFile, Path: "/home/operator/.config/systemd/user/watcher.service"},
@@ -301,102 +283,3 @@ func TestNothingIsReadBeforeADeclaration(t *testing.T) {
t.Fatalf("before a declaration every unit would read as the machine's: %+v", st)
}
}
// Issue 764: a look starts no account's manager. Only an account the declaration places a user-scoped
// unit for is read at all, and its manager only while the machine's says it runs.
func TestAnAccountWithNoUserUnitsIsNeverRead(t *testing.T) {
o := OwnedBy(declared(), nil)
if strings.Join(o.Accounts, " ") != "operator" {
t.Fatalf("only an account with a user-scoped unit is read: %v", o.Accounts)
}
r := shanks()
j := New(r)
j.Set(o)
lookTwice(t, j)
for _, m := range r.read {
if m == "user:telegram" {
t.Fatalf("a module account the declaration only creates had its manager read: %v", r.read)
}
}
}
func TestAManagerNotRunningIsAwayAndNotRead(t *testing.T) {
r := shanks()
j := New(r)
j.Set(OwnedBy(declared(), nil))
lookTwice(t, j)
if len(j.Last().Owned()) != 1 {
t.Fatalf("logged in, the Razer daemon is said: %+v", j.Last().Failed)
}
// The operator logs out: the manager is not asked, nothing in it is failed, it is said away.
r.away = map[string]bool{"operator": true}
r.read = nil
st, changed := j.Look(context.Background())
if strings.Join(r.read, " ") != "system" {
t.Fatalf("a manager not running is not read: %v", r.read)
}
if !changed || len(st.Away) != 1 || !strings.Contains(st.Away[0], "operator") || len(st.Unread) != 0 {
t.Fatalf("away, said, and not unread: %+v", st)
}
for _, f := range st.Failed {
if f.Scope == ScopeUser {
t.Fatalf("a manager not running has no failed units: %+v", f)
}
}
if st.State != Degraded {
t.Fatalf("the machine's own failed mounts still say degraded: %q", st.State)
}
// Back in: read as before, and a failure is new again (two looks).
r.away = nil
st, _ = j.Look(context.Background())
if len(st.Away) != 0 || len(st.Owned()) != 0 {
t.Fatalf("a manager running again is read, its failure on its first look not yet said: %+v", st)
}
st, _ = j.Look(context.Background())
if len(st.Owned()) != 1 {
t.Fatalf("and said on the second: %+v", st.Failed)
}
}
func TestAnAccountTheMachineLacksIsUnread(t *testing.T) {
r := shanks()
r.noAccount = map[string]bool{"operator": true}
j := New(r)
j.Set(OwnedBy(declared(), nil))
st, _ := j.Look(context.Background())
if strings.Join(r.read, " ") != "system" || len(st.Unread) != 1 {
t.Fatalf("an account the machine lacks is unread and its manager never asked: %v %+v", r.read, st)
}
}
func TestWhetherAManagerRunsIsAskedOfTheMachines(t *testing.T) {
var asked []string
state := "inactive"
run := func(_ context.Context, name string, args ...string) (string, error) {
asked = append(asked, name+" "+strings.Join(args, " "))
switch name {
case "getent":
return "telegram:x:10007:10007::/var/lib/telegram:/usr/bin/nologin\n", nil
case "systemctl":
if state != "active" {
return state + "\n", errors.New("systemctl exited 3")
}
return state + "\n", nil
}
return "", errors.New("unexpected")
}
e := Exec{Run: run, System: "arch"}
if running, exists, err := e.UserManagerRunning(context.Background(), "telegram"); running || !exists || err != nil {
t.Fatalf("an inactive manager is not running: %v %v %v", running, exists, err)
}
state = "active"
if running, _, err := e.UserManagerRunning(context.Background(), "telegram"); !running || err != nil {
t.Fatalf("an active manager runs: %v %v", running, err)
}
for _, a := range asked {
if a != "getent passwd telegram" && a != "systemctl is-active user@10007.service" {
t.Errorf("asked something but the user database and the machine's manager: %q", a)
}
}
}
validate/validate_test.go
-13
View File
@@ -18,20 +18,7 @@ func TestTheValidatorIsTheHosts(t *testing.T) {
if problems := Declaration([]byte(ordered)); problems != nil {
t.Fatalf("a declaration carrying its order was refused: %v", problems)
}
// And one carrying a file's owner-at-runtime and the files the provisioner fills (novox/hq ADR 0349 §5): the
// reader takes them before any controller sends them.
owned := `{"declaration": 1, "filled_files": ["/etc/x"], "resources": [{"id": "f", "type": "file", "path": "/etc/x", "content": "a", "owner-at-runtime": "owned"}]}`
if problems := Declaration([]byte(owned)); problems != nil {
t.Fatalf("a declaration saying who owns a file at runtime was refused: %v", problems)
}
// And one naming the provisioner client's account (novox/hq ADR 0349 §4).
named := `{"declaration": 1, "provisioner_account": "mesh-node-provisioner", "filled_files": ["/etc/x"], "resources": [{"id": "f", "type": "file", "path": "/etc/x", "content": "a", "owner-at-runtime": "owned"}]}`
if problems := Declaration([]byte(named)); problems != nil {
t.Fatalf("a declaration naming the provisioner client's account was refused: %v", problems)
}
for _, bad := range []string{
`{"declaration": 1, "provisioner_account": "root", "resources": [{"id": "f", "type": "file", "path": "/etc/x", "content": "a"}]}`,
`{"declaration": 1, "filled_files": ["/etc/none"], "resources": [{"id": "f", "type": "file", "path": "/etc/x", "content": "a", "owner-at-runtime": "owned"}]}`,
`{"declaration": 1, "resources": []}`,
`{"declaration": 99, "resources": [{"id": "d", "type": "directory", "path": "/x"}]}`,
`{"declaration": 1, "resources": [{"id": "d", "type": "nothing-the-host-knows"}]}`,