Commit Graph
54 Commits
Author SHA1 Message Date
jschoubben 35ff67bda0 Design 47's last milestones: the keeper, self-check, healers, hand-acts, provisioner, vault, artifact sweep, disk readings, credential forms (issues 774, 1952, 1953, 1954, 1955)
mesh/check failed: integration: mesh-self-check/cmd/mesh-self-check.TestTheSelfCheckWholeRunsOnTheBus: the run answered {Run:doctor-1793351737-1 At:20…
mesh/delivery-group group feat/774-l3-reland rejected: composed together: the checker's verdict at the members' heads (b48c90fea31f): integration: mesh-self-ch…
mesh/delivery closed: a newer head of the same pull request
One commit with exactly the tree of the reviewed head 35b4509c0a, on main 12c2574c9d, so two compiled programs added and removed on the branch leave no trace in history (issue 1577).
2026-10-30 02:21:51 +01:00
jschoubben c8240c38bb Revert L3 (mesh-catalog, merge 15552413): its node-engine removed every held module on novox, the bus server among them, at 08:54 CET on 2026-10-29 (issue 774)
mesh/delivery merged: source.main moved to novox/mesh-catalog@030f9ee4e780
mesh/delivery-group group revert/774-l3 merged: every member is merged
mesh/check being carried to c8240c38bb89 on 030f9ee4e780 | from 1555241393fe | novox/mesh-catalog#557 in group revert/774-l3
L3 lands again with two fixes: a held module is never removed, and the grants cover L3's new subjects.
2026-10-29 09:42:30 +01:00
jschoubben c193d2b389 Pin mesh-sdk 8eafc0c, the L3 group's head on mesh-sdk main with #120 and #121 (issue 774)
mesh/delivery-group group feat/774-vault-mints-l2 ready: every member ready, and composed together they pass
mesh/delivery closed: a newer head of the same pull request
mesh/check pass c193d2b38970 on 158348ede789 | unit 1319 run, 5401 kept, 47 skipped | integration 1359 run, 5490 kept, 49 skipped | over bound: unit, …
2026-10-28 21:56:06 +01:00
jschoubben 5be6576117 Follow L2 under L3: the vault and mesh-config read the seat mesh-placement's placement, and the check that only listed modules read the controller's assignments moves into noplacement, as L2 deleted the placement check (issue 774, ADR 0567, ADR 0549 §3, §6) 2026-10-28 21:54:36 +01:00
jschoubben 24fad2b92a mesh-vault: pin mesh-sdk 4e6e5cc, the L3 group's head rebased on L2 (issue 774) 2026-10-28 21:54:36 +01:00
jschoubben d3eb5cae80 mesh-vault: secret-ask answers the callers its manifest names, each named as the runtime proved it, within the same bounds; the controller's relay as before (issue 774, ADR 0563 §8) 2026-10-28 21:54:36 +01:00
jschoubben c00005c141 mesh-vault signs: its signing key and authority made once with their operator copies, the verb sign for a node's own certificate from bus state, the event signed, the state signers and the tools that read and reassert them (issue 774, ADR 0550 §2) 2026-10-28 21:54:36 +01:00
jschoubben e6b913cac7 mesh-vault: account_retire refuses while the controller holds the lease or before the wants are read whole, as the passes do (issue 774, review round 1) 2026-10-28 21:54:36 +01:00
jschoubben 411c1b8de8 mesh-vault: prove the vault whole on a real bus mints only from the lease write, and try a failed mint again at the pass; adopt the operator key before the watches open (issue 774, ADR 0561 §1-§4) 2026-10-28 21:54:36 +01:00
jschoubben 64e6688ec8 mesh-vault: tell where modules stand and which node holds the bus from the catalogue's placement first, so the vault does not depend on the controller's assignments for them (issue 774, ADR 0549 §3, §5) 2026-10-28 21:54:36 +01:00
jschoubben 2d57c9663a mesh-vault: serve issue-account, the hand path that mints one wanted account now, and record a rotation at most once per loaded list's revision; a covered account still never rotates in place (issue 774, ADR 0561 §4, §9) 2026-10-28 21:54:36 +01:00
jschoubben 5c1729523b mesh-vault: retire an account the grants part no longer wants only within ADR 0230's bounds, keeping its value, and re-enable it when wanted again; account_retire for a person past the bound (issue 774, ADR 0561 §2) 2026-10-28 21:54:36 +01:00
jschoubben 1628db6c80 mesh-vault: mint only while the controller's lease is handed away, read afresh before each mint, and say in vault_status what is held back and why (issue 774, ADR 0561 §3, §9) 2026-10-28 21:54:36 +01:00
jschoubben 0b9f19b612 mesh-vault: mint each account the grants part wants and the vault holds no value or adopted hash for, in the controller's shape, hashed before account-issued and secret-changed (issue 774, ADR 0561 §1, §2, §4) 2026-10-28 21:54:36 +01:00
jschoubben 59cad7154a Pin mesh-sdk to the group's head 38dbe1d in the build modules, mesh-config, the gate and the vault; the vault's test answers by proposal (issue 1126) 2026-10-28 18:52:35 +01:00
jschoubben a9d96b30be mesh-vault: a watch of the assignments begun again keeps nothing an earlier attempt took, and a module the mirror lacks is read directly before a credential is refused (issue 1126, review round 1)
mesh/check pass a9d96b30be8d on 69056948bfa8 | unit 1639 run, 4201 kept, 46 skipped | integration 1674 run, 4215 kept, 47 skipped | over bound: unit, …
mesh/delivery merged: source.main moved to novox/mesh-catalog@85bbeb37e3e0
mesh/delivery-group group fix/1126-vault-own merged: every member is merged
A watch that failed partway kept what it had taken; begun again, the runtime does not hand a key deleted meanwhile, so
a removed assignment read as standing and credentialRefusal could hand a consumer credential to a node no longer
running its module. Each attempt now begins from nothing held, and a late delivery to an older attempt changes nothing.
A 'not on the node' refusal is confirmed by a direct read first, since a watched mirror can be behind.
2026-10-28 14:55:36 +01:00
jschoubben 277a4d0060 mesh-vault answers own in one pass: assignments from its watch, each module's once, one synced record per answer (issue 1126)
Seed-9's probe (13:31-13:36Z): novox's own of 57 secrets and 146 credentials took 26 s, past the node-engine's 15 s,
and ace and novox refused their whole shadow answer. own read the controller's assignment over the bus once per secret
and per credential (203 reads), read every record of every node, and synced the hand-out record once per value, under
the one lock every node's answers share.

Now the vault watches the assignments mirror (a watch, never a poll) and reads it from memory once caught up, over the
bus until then; one answer reads each module's assignment once; only own/<node>/ records are read; and every hand-out
of one answer is recorded in one synced write, none handed when it cannot be.

Quick fix weighed: a longer bound in the node-engine (VaultWithin). Thorough fix chosen: the cost grows with every
module and every node, and the lock is shared by every shadow answer and every apply.
2026-10-28 14:47:20 +01:00
jschoubben 0974d1a57d mesh-vault's own_drift row carries the id as a node runs it, its module's name first, and the vault pins mesh-sdk 3028c7e (issue 1126, cause C4; review round 2)
mesh/check being carried to 0974d1a57d99 on 2a03c106f4b9 | from 9fd95693db54
mesh/delivery merged: source.main moved to novox/mesh-catalog@9fd95693db54
mesh/delivery-group group fix/1126-seed-same merged: every member is merged
2026-10-28 10:02:46 +01:00
jschoubben a17356a162 Fill go.sum for mesh-sdk 1062b33, the group's head (issue 1126) 2026-10-28 09:53:34 +01:00
jschoubben 35ad58e6e2 mesh-vault pins the group's mesh-sdk head 1062b33 (issue 1126; go.sum is filled once that head is pushed) 2026-10-28 09:51:48 +01:00
jschoubben da5da4f6f7 The vault takes what runs through move-in's replace_running, not a verb of its own, so no seat promises anything new (issue 1126, cause C4) 2026-10-28 09:48:14 +01:00
jschoubben 671a6a1f47 own-take-running has a person's switch, own_take_switch, kept in the vault's store (issue 1126, cause C4) 2026-10-28 09:39:16 +01:00
jschoubben 47ed384d97 The vault takes the value that runs on a machine over the one it was first handed, from that machine alone (issue 1126, cause C4)
move-in never replaces a value the vault holds, so where the controller made an own secret anew after the move, the
new value runs while the vault keeps the first. The verb own-take-running takes it from the node-engine of the machine
it is for, keeps the replaced value as the previous generation, raises the generation and says own-replaced without a
value; own_drift lists every own secret a node's shadow answer says another value of, mesh-wide. The module needs the
mesh-sdk head of fix/1126-seed-same-own-resync (delivery.OwnValuesDiffer, the seat's verb).
2026-10-28 09:37:54 +01:00
jschoubben 1faf317069 Review of L2: the vault's own fails closed: a pair credential's consumer is read standing, the shared value goes only to a holder its record names, and unread assignments refuse, never hand (C1, C2); the placement check pins mesh-sdk 3968849 (issue 1126, ADR 0550 §5)
mesh/delivery closed: a newer head of the same pull request
mesh/check checking 1faf31706919 on 5b75233406fe | novox/mesh-catalog#505 in group feat/1126-nodes-compose
2026-10-26 20:12:21 +01:00
jschoubben 851e8782d5 Nodes compose: the catalogue's placement files and their check, mesh-config reading placement, the vault's own and pair credentials, the forge's commit trees (issue 1126)
mesh/check checking 851e8782d569 on da29f77b1081 | novox/mesh-catalog#505 in group feat/1126-nodes-compose
mesh/delivery closed: a newer head of the same pull request
One commit for the branch feat/1126-nodes-compose of mesh-catalog: its 15 commits squashed so its history does not carry two compiled programs committed by mistake and removed later, mesh-config (5.5 MB) and checks/placement (6.2 MB) (issue 1577). The tree equals the former head 2db56067. The operator allowed this rewrite of the feature branch on 2026-10-26.
2026-10-26 19:49:39 +01:00
jschoubben c9414b1c98 Vault: pin mesh-sdk at the switch group's head c2a410bb, one sdk for the adoption and the switch (issue 1945)
mesh/delivery merged: source.main moved to novox/mesh-catalog@879600202a09
mesh/delivery-group group feat/1945-user-list-switch merged: every member is merged
mesh/check checking c9414b1c9856 on 879600202a09 | novox/mesh-catalog#494 in group feat/1945-user-list-switch
2026-10-26 12:31:56 +01:00
jschoubben e2136ec0c9 The vault takes a handing whole once, refusals and all, and names the refused users in its last adoption (review of the adoption, ADR 0547 clause 9) 2026-10-26 12:03:14 +01:00
jschoubben f0fd53281b The vault says once why a look of the bus module's have adopted nothing, and again when it changes (ADR 0547 clause 9) 2026-10-26 11:37:14 +01:00
jschoubben fb039250f6 The vault adopts the hashes of the bus accounts it holds no value for from the bus node's handing (ADR 0547 clause 9, issue 1945)
take-over waits on a hash for every user of the loaded list, and the vault holds no value for the controller's own
account, the node accounts and the enrolled devices. The bus node's engine now hands every loaded user's hash in the
bus module's have, sealed to the vault's key (mesh-host and mesh-sdk on the same branch). The vault answers its key in
node-key, reads the haves (manifest reads node-engine.have, granted from the manifest, no controller change), watches
the bus module's have, opens the handing with its own key, and writes each user it holds no value for into
bus-accounts marked adopted, with the list and node it came from, saying account-hashed. It never touches a user whose
covered account it holds or whose key is its own hash, writes an adopted key again only for another hash, and refuses
a handing from any node the assignments do not name as holding the bus module.

Tools: account_adopt adopts now from the have as the watch does; secret_own, unnarrowed by module, lists the adopted
accounts and the last adoption.
2026-10-26 11:36:33 +01:00
jschoubben e7aaa8420e Vault: name a hash record's secret in a field called of, so the runtime accepts it
mesh/delivery merged: source.main moved to novox/mesh-catalog@3ded9e780043
mesh/check pass e7aaa8420eac on 3ded9e780043 | unit 2018 run, 2325 kept, 5 skipped | integration 2040 run, 2344 kept, 9 skipped | over bound: unit, in…
The runtime refuses a state field whose name ends in secret, so no bus-account hash could be written on novox. The test state now applies the same check.

Issue: 1945
2026-10-26 09:25:35 +01:00
jschoubben 1c55e5d44d Vault: take and hash a process's own bus account (broker-<process id>) as a covered account, so take-over is not held by every process user
mesh/delivery merged: source.main moved to novox/mesh-catalog@22381d75224f
mesh/check failed: integration: meshbuildnode/cmd/mesh-build-node.TestAFixtureBundleBuiltTwiceHasOneDigest: no toolchain of this lock is installed und…
The controller hands each process account of a core or moved module with origin issued (ADR 0436 §2); the vault took
only broker and provisioner-broker, refused these and never hashed them. Owed hashes are now said once per pass, and
account_hash says who asked when it refuses too (review round 1).

Issue: 1945
2026-10-26 02:35:57 +01:00
jschoubben 32b912cfe5 Vault: hash each bus account it is handed into bus-accounts.<user>, once per generation, so the grants part can compose the user list without the controller
The controller still mints each covered account and hands it to the vault (ADR 0374); ADR 0547 clause 9 has the vault
write the bcrypt hash of what it holds into its own state, readable by the grants part, and say account-hashed. It
hashes on keep, at start every held account with no key of its generation, and again by hand (account_hash); a refused
write is owed and written at the next pass. secret_own says per bus account whether this generation's hash is written.

Issue: 1945
2026-10-26 02:28:28 +01:00
jschoubben e572937d97 Vault: secret_own reads the module-own secrets it keeps, as fingerprints, so a generation can be told without the value
mesh/delivery closed: a newer head of the same pull request
mesh/check pass e572937d97fa on 567736f7795a | unit 2143 run, 2198 kept, 5 skipped | integration 2165 run, 2217 kept, 9 skipped | over bound: unit, in…
Issue: 1945
2026-10-26 02:09:03 +01:00
jschoubben 92c5e1d9ae Pin mesh-provisioner, mesh-vault and rofi to a commit of mesh-sdk main, since the pinned one is on no branch and no fetch can resolve it (issue 1659)
mesh/merge-gate head 92c5e1d9aea8 merged into trunk 38ac738d8463 — pass: builds mesh-provisioner, mesh-vault, rofi → g14, novox, shanks; no bus step; every…
mesh/repo-check head 92c5e1d9aea8 merged into trunk 38ac738d8463 — pass: its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-…
mesh/delivery delivered
2026-10-21 13:14:57 +02:00
jschoubben 60a3823f0e mesh-vault reads an ended assignment as a removed key (issue 1098) 2026-10-18 02:37:28 +02:00
mesh-tools 652b00b25b Say on every directory how it treats files it does not know, and check it (issue 1024, ADR 0422)
mesh/merge-gate checking again: the gate refused `unknown` because the controller judged with a node-engine validator pinned before mesh-host #117; mesh-co…
mesh/repo-check checking again: the gate refused `unknown` because the controller judged with a node-engine validator pinned before mesh-host #117; mesh-co…
mesh/delivery-group group feat/owned-paths checking: 0 of 2 member(s) ready
mesh/delivery superseded: a newer head of the same pull request
Every directory and archive resource carries `unknown`: move-out for a home folder whose program writes nothing there (the node-engine, delivered in mesh-host #117, moves the rest to ~/.local/state/mesh-moved/<module>/ with a list), module for ~/.ssh and ~/.claude, program elsewhere; `knows` keeps what a program or the operator relies on (powerlevel10k's compiled files, i3's scripts and the asus/g14 drop-ins, dunst's slack rule, betterbird's records, node-env's forge credential, issue 1033). Two shared folders are no longer declared. Each module's README says it. checks/owned holds every manifest to it; checks/sshdir stays while main's merge check runs it (issue 1035). Tests that pin a manifest follow the new key only. Reviewed in three rounds.
2026-10-17 16:23:03 +00:00
fix-vault-pas@novox 5a00596908 Make the vault ask a software-issuing provider's tool itself for holders that hold nothing (issue 928) 2026-10-16 19:34:11 +02:00
fix-vault-pas@novox 0dab4475cd Answer an unreadable assignment as not yet, never a refusal (issue 928) 2026-10-16 19:30:15 +02:00
mesh-tools 0260ae92f6 Feature 57 (config and provisioning): release the catalogue, the reviewed release branch merged with main, on top of main 2026-10-16 09:42:45 +00:00
jochen dd14526870 Refuse a daemon's line in what a module shows: the operator types nox (hq ADR 0272, issue 356)
mesh/merge-gate pass: builds mesh-vault → novox; no bus step; every machine composes with the change as it did without (4 of 4 compose)
mesh/repo-check pass: its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)
mesh/delivery covered after its failure: a later walk that contains it was delivered: plan-1792148283013560219 (novox/mesh-catalog at 3db85a0d into main,…
A condition told the operator to type the node-engine's binary at the machine, and nothing held the catalogue's texts to ADR 0272. The words check gains retired-lines — the controller's and the node-engine's commands, copied from hq's words.py --list lines, which compares it — and refuses one in any string a module's tools can show or any manifest description. The vault's and the route adapter's texts now say the nox line.
2026-10-09 18:33:49 +02:00
jochen 08265a70ca audit-logger, model-usage: retry a failed write and never lose the event (hq issue 276)
Both caught a failed write and took the event, losing it silently; the SDK's rule is to throw when
the work was not done. A failed write now throws so the bus offers the event again, and is spooled
on disk at once; on its last delivery the spooled event is taken, and a background pass replays the
spool once writing works. The runtime does not pass the delivery count, so the spool counts failed
deliveries itself, across restarts. Over its bound (1000 events or 30 minutes) the last delivery is
no longer taken, so the bus gives it up and the controller raises max-deliveries - the one existing
condition that names a consumer which cannot keep up - while the spool still holds it.

Writes are idempotent by event: the trail skips an id it already wrote; the usage upsert keeps the
reading observed latest (migration 2), so a late replay never overwrites a newer one. Each module has
a status tool for the spool, declared as valuable data (ADR 0233). model-usage moves to the bundle
shape (ADR 0198) with its schema in a prepare step and numbered migrations; its old container shape
had no image. Both on mesh-sdk 0.1.13.

The log-only handlers of redis, mssql, mosquitto, mongodb, mesh-vault, showcase and the catalogue no
longer throw a TypeError on an event without a body.
2026-10-06 18:36:16 +02:00
jochen 685cb1cb1b Declare every module's data; the backup holder measures it (hq ADR 0233)
Backup lines are derived from each module's data section instead of written by hand; the holder
measures declared items, reads the array under them, and deletes a retired item only after a last
restore point; the Go providers say each held consumer's size so an empty replacement is seen.
2026-10-06 16:47:49 +02:00
jochen 4769dadf63 State each provision's identity bound (hq issue 263)
Consumers were held to an S3 access key's 20 characters whatever they
required. Each provider now says what its backend keeps: minio 20,
PostgreSQL and MongoDB and DNS 63, Gitea 40, a mailbox 64, SQL Server 128,
Keycloak 255, unbounded where the store has no limit, and none for the
resolver and route provisions, which keep no name of their consumers.
Needs the controller that reads the field (mesh-controller, ADR 0225).
2026-10-06 02:16:27 +02:00
jschoubben 32cd92baeb Back up every store: the restic module holds node-backup, the stores contribute their dumps
ADR 0214 / to-be 43. restic keeps one repository per machine and takes a nightly snapshot per
module — 14 daily, 8 weekly, 6 monthly — and restores beside the live data, never over it. postgres,
mssql and mongodb contribute a consistent dump; minio, influxdb, the vault, mailu, gitea and
nextcloud the directories that hold their data.
2026-10-05 11:47:59 +02:00
jochen 043ae17fbf mesh-vault: its handlers, tools and provisioner run in the node's runtime (hq ADR 0198)
The mesh-vault container goes with its Dockerfile, build bases, bus credential and state directory; its env was already host paths, so it becomes the bundle's words unchanged.
2026-10-04 00:52:31 +02:00
jschoubben 89e0dde9e0 The vault's claim and its own event names return
The uplink branch was split from the vault's with the vault's files reset to a main that did not yet
hold #205; merging it afterwards took the older vault definition along (no claim, the refused event
names), and the vault could not be built. Restored to #205's state.
2026-10-01 17:19:05 +02:00
jschoubben 2e96d2f67d This branch carries the uplink capabilities alone (hq ADR 0161 rule 3); merges once every machine running a holder has reported uplink-<manager> 2026-10-01 16:52:47 +02:00
jschoubben 6ba61a8b4b The provisioner announces the vault's events by their new names 2026-10-01 16:51:13 +02:00
jschoubben 8368697744 The vault's events are its own: provisioned, rotated, deprovisioned
A module publishes under its own name only; secret.provisioned read as another module's event and the
builder refused the vault's definition today, so the seat claim could not be built. The three events lose
the prefix; nothing outside the vault listens for the old names.
2026-10-01 16:50:56 +02:00
jschoubben 966fed1829 The vault claims mesh-vault, and each uplink holder declares the manager it speaks for (hq ADR 0161)
The vault's claim makes a second provider of secret a second claimant, refused by name. The three
uplink definitions declare uplink-networkmanager, uplink-systemd-networkd and uplink-dhcpcd, which the
host reports for the manager it finds active, so the holder for a manager the machine does not run
is refused the way any missing capability is. Merges after the controller holds the seat and the host
reports the capability.
2026-10-01 15:58:05 +02:00