Compare commits

...
Author SHA1 Message Date
mesh-admin 4bf705fea7 Merge pull request 'mssql: place its directories and name the software's port, not a machine's' (#160) from feat/mssql-for-ace into main 2026-09-30 13:55:25 +00:00
mesh-admin 00893d4944 Merge pull request 'letta: placed state, a route, accepted keys, and a runtime that can log in' (#171) from feat/letta-for-ace into main 2026-09-30 13:55:23 +00:00
jschoubben a2b9a9a411 Merge pull request 'dnsmasq: pass the DNSSEC bit down from the validating upstreams' (#179) from fix/resolver-passes-the-dnssec-bit into main 2026-09-30 13:17:38 +00:00
jschoubben 9523105df4 dnsmasq: pass the DNSSEC bit down from the validating upstreams
A program that checks its resolver validates — Mailu's admin does, at
start — could not use the mesh's resolver, and the one it ships instead
knows no mesh name (novox/hq issue 171). proxy-dnssec copies the AD bit
from 1.1.1.1 and 8.8.8.8, both of which validate.
2026-09-30 15:17:34 +02:00
jschoubben 4449f44cf1 Merge pull request 'mailu: admin asks the machine's resolver, not Mailu's own' (#178) from fix/mailu-admin-asks-the-machines-resolver into main 2026-09-30 13:13:29 +00:00
jschoubben 47f6e7d78b mailu: admin asks the machine's resolver, not Mailu's own
admin is the one container here that reaches something by a mesh name:
the database, bound in database.env. A mesh name is answered by the
machine's resolver, which the runtime hands every container that does
not name its own (novox/hq ADR 0148); Mailu's unbound knows no mesh
name, and since the mesh stopped copying names into containers admin
could not find its database. The others keep unbound — rspamd needs a
validating resolver for DNSBL lookups and asks for no mesh name.
2026-09-30 15:13:25 +02:00
jschoubben 7ab522ba31 Merge pull request 'dnsmasq: answer a container's query, which arrives on the runtime's bridge' (#176) from fix/110-the-resolver-answers-a-container into main 2026-09-30 12:31:30 +00:00
jschoubben 204bfbbaf9 dnsmasq: answer a container's query, which arrives on the runtime's bridge
dnsmasq admits a query by the interface it arrives on when told interface=,
and by the address it is sent to when told listen-address=. A container's
query is sent to the machine's private address but arrives on docker0, so
interface=mesh0 dropped it silently on every machine (novox/hq issue 110).
Name the address, not the interface.
2026-09-30 14:31:26 +02:00
jschoubben e0c09f46b6 Merge pull request 'dnsmasq: the runtime is reloaded when its dns file changes, and may then be restarted safely' (#175) from fix/110-the-runtime-reads-its-dns into main 2026-09-30 12:22:10 +00:00
jschoubben e0faf012be dnsmasq: the runtime is reloaded when its dns file changes, and may then be restarted safely
novox/hq 04-ISSUES/110. The module writes the runtime's `dns` key and
deliberately ordered no restart, because a restart stops every container.
It also ordered no reload, and the key holds only for containers created
after the runtime next starts. On two of four machines the runtime
predated the file — one since August — so every container there was
handed a public resolver and no mesh name resolved, while everything read
as fine. The third machine works only because its runtime happened to
restart later.

The file now also sets live-restore, which the runtime reads on a reload,
and the module declares the runtime reloaded when the file changes (ADR
0102: reload, don't restart). A reload still does not make `dns` take
effect; what it does is make the one restart that key needs keep every
container running. That restart stays the operator's, once per machine,
and is harmless from the second time on.

The mesh restarts nothing here. Undeclared, the runtime's unit goes back
to the state it was found in (ADR 0118).
2026-09-30 14:22:03 +02:00
mesh-admin 1b19c79d63 Merge pull request 'postgres: the provisioner dials the port the mesh gave the store' (#174) from fix/postgres-provisioner-dials-the-port-it-was-given into main 2026-09-30 12:06:25 +00:00
jschoubben 4ec2ae1f7f postgres: the provisioner dials the port the mesh gave the store
MESH_PROVISION_POSTGRES named 127.0.0.1:5432 literally; the seat twin
(${seat:mesh-store:5432}) corrected it only on the machine holding the
mesh-store seat. On any other machine — ace, where the module provides
postgres-database without the seat — the twin is empty and the sidecar
would have dialled whatever else holds 5432 (HAL's postgres). ${port:5432}
is the machine port on every node; on novox the twin still says the same
6852, so nothing moves there.
2026-09-30 14:01:53 +02:00
jschoubben 8f459c7023 letta: placed state, a route, accepted keys, and a runtime that can log in
The module named /var/lib/letta, a layout no definition may carry (ADR
0112); state is now a placed directory holding the bindings and secrets.

letta had no route, while the server it replaces is reached by its public
name (a workflow calls it there). It now contributes one for its `web`
endpoint; reach is the assignment's.

The server needs an OpenAI key for agents on OpenAI models, and nothing
gave it one: `openai-api-key` is an own-secret, accepted from the
operator (a key someone chose, not one the mesh can mint). The
server-password is accepted the same way where a server already has
clients. Both, and the database password inside LETTA_PG_URI, stay in the
server's environment: letta 0.6.x reads settings from the environment
only, and its startup.sh starts an embedded PostgreSQL unless
LETTA_PG_URI is set - the declared reason now says so.

The runtime's tools never authenticated: the client sent only a Bearer
token, and 0.6.x's --secure mode checks X-BARE-PASSWORD ("password <it>")
and answers 401 otherwise. The client now sends both. Its password comes
from the runtime config file (the key client.ts reads first) instead of an
env-file, so the runtime container no longer carries a secret in its
environment.

Image: the same 0.6.8 image, now pinned by the index digest ace runs
rather than its amd64 manifest.

Verified: catalogue tests with MESH_CATALOGUE set; tsc -p tsconfig.json in
the mesh-tools build image. Throwaway containers: a fresh 0.6.8 with its
embedded PG and two blocks made through the API; stopped, copied, dumped
from the copy; restored (schema letta + vector pre-made by the superuser,
--no-owner --role, search_path set on the database as the original had
it) into a grant-shaped database on the postgres module's pgvector image
(PG17); started in this shape: alembic finds nothing to do, both blocks
are there, a wrong password is refused with 401, and the patched client
lists agents. Test containers and data removed.
2026-09-30 12:22:12 +02:00
jschoubben 0fce3ebf5d mssql: place its directories and name the software's port, not a machine's
The manifest stated /var/lib/mssql, its grants and its SA file by path, and
declared it listens on 4848 - the port one machine's predecessor published,
which is an assignment's fact (ADR 0112, 0138). ace is moving its own
SQL Server (80 GB of work databases) onto the mesh, so the module has to be
the same on every machine.

- state is the assignment's root (place "."), grants is placed, and every
  reference (sa.env, the env-file, the SA and grants mounts, receives,
  grants, own-secrets) names them as ${dir:...}.
- the database endpoint listens on 1433, the port SQL Server uses; a machine
  that must keep an older number pins it in its assignment.
- the image pin is unchanged: it is the digest ace runs today (CU27,
  16.0.4295), the same as novox.

novox is untouched: rendered with novox's own setting ({"ports":{"1433":4848}})
through the controller's Declaration and Rules, every resource - paths,
container names, volumes, env-file, the 4848:1433 mapping, owners, modes - is
byte-identical to what main renders; the only difference is the firewall
rule's comment text (still port 4848, from the mesh).

Verified: catalogue tests pass with MESH_CATALOGUE on this tree. The pinned
image ran as a throwaway on a 0700 10001:0 data dir with a root-owned 0600
env-file (dummy SA), answered sqlcmd as sa; a scratch database stopped,
copied with cp -a, checksummed and started on the copy kept its rows and
CHECKSUM_AGG.
2026-09-30 11:57:45 +02:00
6 changed files with 56 additions and 51 deletions
File diff suppressed because one or more lines are too long
+8 -4
View File
@@ -1,10 +1,10 @@
// The Letta API client — letta's own code, living in the module (novox/hq ADR 0039). Its tools
// import it; nothing outside letta does.
//
// Letta authenticates with a single server password, presented as a Bearer token. That password is
// a mesh own-secret, minted once and handed to both the server (LETTA_SERVER_PASSWORD) and this
// client (MESH_LETTA_PASSWORD) — so the module's tools are live without anything configured by hand.
// The runtime config file may still override the URL or password.
// Letta authenticates with a single server password. That password is a mesh own-secret handed to
// both the server (LETTA_SERVER_PASSWORD) and this client, through the runtime config file the mesh
// mounts (its `password` key) — so the module's tools are live without anything configured by hand.
// Where a server already has clients, the password is accepted rather than minted.
import { readFileSync } from "node:fs";
@@ -58,6 +58,10 @@ export class LettaClient {
...options,
headers: {
"Content-Type": "application/json",
// The server's --secure mode checks X-BARE-PASSWORD ("password <it>") and answers a Bearer
// token alone with 401 (letta/server/rest_api/app.py, 0.6.x). Both are sent: Bearer is what
// later servers read.
"X-BARE-PASSWORD": `password ${this.password}`,
Authorization: `Bearer ${this.password}`,
...(options.headers as Record<string, string> | undefined),
},
+21 -25
View File
@@ -5,21 +5,28 @@
"container-runtime"
],
"requires": [
"postgres-database"
"postgres-database",
"route"
],
"contributes": {
"postgres-database": {
"name": "letta"
},
"route": {
"label": "letta",
"endpoint": "web"
}
},
"binds": {
"postgres-database": "/var/lib/letta/database.json"
"postgres-database": "${dir:state}/database.json",
"route": "${dir:state}/route.json"
},
"secrets": {
"postgres-database": "/var/lib/letta/database.secret"
"postgres-database": "${dir:state}/database.secret"
},
"own-secrets": {
"server-password": "/var/lib/letta/server-password.secret",
"server-password": "${dir:state}/server-password.secret",
"openai-api-key": "${dir:state}/openai-api-key.secret",
"broker": "/var/lib/mesh/letta/broker"
},
"listens": [
@@ -28,7 +35,7 @@
"port": 8283,
"protocol": "tcp",
"from": "mesh",
"why": "the Letta agent server REST API and web UI; a public name is a route grant later"
"why": "the Letta agent server REST API and web UI, password-protected (--secure); a public name is the route's"
}
],
"resources": [
@@ -41,15 +48,15 @@
{
"id": "state",
"type": "directory",
"path": "/var/lib/letta",
"mode": "0700"
"mode": "0700",
"place": "."
},
{
"id": "server-env",
"type": "file",
"path": "/var/lib/letta/server.env",
"path": "${dir:state}/server.env",
"mode": "0600",
"content": "LETTA_PG_URI=postgresql://${bound:postgres-database:as}:${secret:postgres-database}@${bound:postgres-database:at}:${bound:postgres-database:port}/${bound:postgres-database:as}\nLETTA_SERVER_PASSWORD=${secret:server-password}\nSECURE=true\nTZ=Europe/Brussels\n"
"content": "LETTA_PG_URI=postgresql://${bound:postgres-database:as}:${secret:postgres-database}@${bound:postgres-database:at}:${bound:postgres-database:port}/${bound:postgres-database:as}\nLETTA_SERVER_PASSWORD=${secret:server-password}\nOPENAI_API_KEY=${secret:openai-api-key}\nSECURE=true\nTZ=Europe/Brussels\n"
},
{
"id": "net",
@@ -60,31 +67,24 @@
"id": "server",
"type": "container",
"name": "letta",
"image": "letta/letta@sha256:1d2e0692514287c5ed1a483e14e16ed945f8632d315539f5e66373bb7d7c471b",
"image": "letta/letta@sha256:bfd1e49ce45b9a208c941e832c1d1d194017ff210a3784b0ca6c323aed767a29",
"network": "letta",
"env-file": [
"/var/lib/letta/server.env"
"${dir:state}/server.env"
],
"ports": [
"8283"
],
"secrets-in-environment": "the letta image is env-driven and its file-source support could not be verified; the mesh runtime can take its password from config.json (client.ts) \u2014 not yet converted"
"secrets-in-environment": "letta 0.6.x reads its settings from the environment only (pydantic settings, no secrets_dir or _FILE twin), and its startup.sh starts an embedded PostgreSQL unless LETTA_PG_URI is set - so the database password travels inside that URI (startup.sh also echoes it to the log); LETTA_SERVER_PASSWORD and OPENAI_API_KEY have no file source either"
},
{
"id": "runtime-config",
"type": "file",
"path": "/var/lib/mesh/letta/config.json",
"mode": "0600",
"content": "{}\n",
"content": "{\n \"password\": \"${secret:server-password}\"\n}\n",
"merge": "json"
},
{
"id": "runtime-env",
"type": "file",
"path": "/var/lib/letta/runtime.env",
"mode": "0600",
"content": "MESH_LETTA_PASSWORD=${secret:server-password}\n"
},
{
"id": "runtime",
"type": "container",
@@ -99,14 +99,10 @@
"MESH_LETTA_URL": "http://letta:8283",
"MESH_LETTA_CONFIG_FILE": "/run/config/config.json"
},
"env-file": [
"/var/lib/letta/runtime.env"
],
"restart-on": [
"runtime-config"
],
"artifact": "runtime",
"secrets-in-environment": "the letta image is env-driven and its file-source support could not be verified; the mesh runtime can take its password from config.json (client.ts) \u2014 not yet converted"
"artifact": "runtime"
}
],
"build": {
+2 -5
View File
@@ -120,7 +120,7 @@
"port": 7080,
"protocol": "tcp",
"from": "mesh",
"why": "the web front over http; only the ACME HTTP-01 passthrough is routed here \u2014 everything else 301s to https and would loop a proxy"
"why": "the web front over http; only the ACME HTTP-01 passthrough is routed here — everything else 301s to https and would loop a proxy"
},
{
"name": "web-tls",
@@ -315,10 +315,7 @@
"${dir:data-data}:/data",
"${dir:data-dkim}:/dkim"
],
"secrets-in-environment": "mailu-admin honours SECRET_KEY_FILE, DB_PW_FILE and API_TOKEN_FILE (configuration.py) but INITIAL_ADMIN_PW is env-only (start.py); the remaining containers' need for SECRET_KEY is unverified",
"dns": [
"192.168.203.254"
]
"secrets-in-environment": "mailu-admin honours SECRET_KEY_FILE, DB_PW_FILE and API_TOKEN_FILE (configuration.py) but INITIAL_ADMIN_PW is env-only (start.py); the remaining containers' need for SECRET_KEY is unverified"
},
{
"id": "imap",
+11 -12
View File
@@ -21,23 +21,23 @@
"listens": [
{
"name": "database",
"port": 4848,
"port": 1433,
"protocol": "tcp",
"from": "mesh",
"why": "modules on any machine that were granted a database. 4848, not 1433: the machine this replaces has served it there since it was installed, and every consumer was handed that number"
"why": "modules on any machine that were granted a database, and the people who were given its address; the machine port is the assignment's to pin"
}
],
"serves": {
"mssql-database": {}
},
"receives": {
"mssql-database": "/var/lib/mssql/grants/mesh.json"
"mssql-database": "${dir:grants}/mesh.json"
},
"grants": {
"mssql-database": "/var/lib/mssql/grants"
"mssql-database": "${dir:grants}"
},
"own-secrets": {
"sa": "/var/lib/mssql/sa.secret",
"sa": "${dir:state}/sa.secret",
"broker": "/var/lib/mesh/mssql/broker"
},
"resources": [
@@ -50,19 +50,18 @@
{
"id": "state",
"type": "directory",
"path": "/var/lib/mssql",
"mode": "0700"
"mode": "0700",
"place": "."
},
{
"id": "grants",
"type": "directory",
"path": "/var/lib/mssql/grants",
"mode": "0700"
},
{
"id": "sa-env",
"type": "file",
"path": "/var/lib/mssql/sa.env",
"path": "${dir:state}/sa.env",
"mode": "0600",
"content": "ACCEPT_EULA=Y\nMSSQL_SA_PASSWORD=${secret:sa}\n"
},
@@ -84,7 +83,7 @@
"image": "mcr.microsoft.com/mssql/server@sha256:4402d880dd4c34bfa7d8705e56a86cd6c88da80a1f6bbbe741f999e76264a090",
"network": "mssql",
"env-file": [
"/var/lib/mssql/sa.env"
"${dir:state}/sa.env"
],
"ports": [
"1433"
@@ -101,8 +100,8 @@
"network": "mssql",
"volumes": [
"/var/lib/mesh/mssql/broker:/run/secrets/broker:ro",
"/var/lib/mssql/grants:/var/lib/mssql/grants:ro",
"/var/lib/mssql/sa.secret:/run/secrets/sa:ro"
"${dir:grants}:/var/lib/mssql/grants:ro",
"${dir:state}/sa.secret:/run/secrets/sa:ro"
],
"env": {
"MESH_PROVISION_MSSQL": "mssql://sa@mssql:1433/master",
+1 -1
View File
@@ -105,7 +105,7 @@
"/var/lib/postgres/superuser.secret:/run/secrets/superuser:ro"
],
"env": {
"MESH_PROVISION_POSTGRES": "postgres://postgres@127.0.0.1:5432/postgres?sslmode=disable",
"MESH_PROVISION_POSTGRES": "postgres://postgres@127.0.0.1:${port:5432}/postgres?sslmode=disable",
"MESH_PROVISION_POSTGRES_PORT": "${seat:mesh-store:5432}",
"MESH_PROVISION_PASSWORD_FILE": "/run/secrets/superuser",
"MESH_BROKER_FILE": "/run/secrets/broker",