migrateLegacySchedules (internal/app/archied/state_store.go:191) replaces the ENTIRE legacy cron/jobs.json list into the schedules resource via control.Command, but validateSchedules (internal/app/controlplane/schedules.go:29) refuses every job whose Kind is not exactly 'workflow' (cronstore.KindChat='chat' is the legacy default; an empty kind on schema-v1 jobs means chat too). A deployment whose legacy cron store holds one chat or empty-kind job therefore fails Command with ErrValidation, migrateLegacySchedules returns the error, and RunStateStore aborts ('migrate schedules: ...') before the server ever listens. The failure is permanent: ImportConfig has already seeded schedules at version 1, so on every subsequent restart versions[controlplane.SchedulesKind]==1 makes the migration re-run and fail again. The only remedy is manually renaming cron/jobs.json, which no recovery doc mentions. The same hard-fail applies to an existing-but-unreadable or corrupt jobs.json (os.Stat succeeds, cronstore.Open/List errors are returned fatal). Legacy schedules were previously tolerated as a live concern (the delivery router refuses unknown kinds per-job); the migration should skip or quarantine refused jobs rather than fail the whole State Store. Docs: docs/prds/runtime-control-plane.md 'a stale TOML value ... must not stop this process from starting' — the same spirit the seed path honours via SeedSkip.
Close reason
Closed on GitHub; syncing bead state
Imported from beads archie-core-q6vw · GitHub #1046 · type bug · priority P1 · created 2026-09-21 · closed 2026-09-26
migrateLegacySchedules (internal/app/archied/state_store.go:191) replaces the ENTIRE legacy cron/jobs.json list into the schedules resource via control.Command, but validateSchedules (internal/app/controlplane/schedules.go:29) refuses every job whose Kind is not exactly 'workflow' (cronstore.KindChat='chat' is the legacy default; an empty kind on schema-v1 jobs means chat too). A deployment whose legacy cron store holds one chat or empty-kind job therefore fails Command with ErrValidation, migrateLegacySchedules returns the error, and RunStateStore aborts ('migrate schedules: ...') before the server ever listens. The failure is permanent: ImportConfig has already seeded schedules at version 1, so on every subsequent restart versions[controlplane.SchedulesKind]==1 makes the migration re-run and fail again. The only remedy is manually renaming cron/jobs.json, which no recovery doc mentions. The same hard-fail applies to an existing-but-unreadable or corrupt jobs.json (os.Stat succeeds, cronstore.Open/List errors are returned fatal). Legacy schedules were previously tolerated as a live concern (the delivery router refuses unknown kinds per-job); the migration should skip or quarantine refused jobs rather than fail the whole State Store. Docs: docs/prds/runtime-control-plane.md 'a stale TOML value ... must not stop this process from starting' — the same spirit the seed path honours via SeedSkip.
### Close reason
Closed on GitHub; syncing bead state
---
_Imported from beads `archie-core-q6vw` · GitHub #1046 · type bug · priority P1 · created 2026-09-21 · closed 2026-09-26_
sam
added the highbug labels 2026-10-27 13:40:45 +00:00
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
migrateLegacySchedules (internal/app/archied/state_store.go:191) replaces the ENTIRE legacy cron/jobs.json list into the schedules resource via control.Command, but validateSchedules (internal/app/controlplane/schedules.go:29) refuses every job whose Kind is not exactly 'workflow' (cronstore.KindChat='chat' is the legacy default; an empty kind on schema-v1 jobs means chat too). A deployment whose legacy cron store holds one chat or empty-kind job therefore fails Command with ErrValidation, migrateLegacySchedules returns the error, and RunStateStore aborts ('migrate schedules: ...') before the server ever listens. The failure is permanent: ImportConfig has already seeded schedules at version 1, so on every subsequent restart versions[controlplane.SchedulesKind]==1 makes the migration re-run and fail again. The only remedy is manually renaming cron/jobs.json, which no recovery doc mentions. The same hard-fail applies to an existing-but-unreadable or corrupt jobs.json (os.Stat succeeds, cronstore.Open/List errors are returned fatal). Legacy schedules were previously tolerated as a live concern (the delivery router refuses unknown kinds per-job); the migration should skip or quarantine refused jobs rather than fail the whole State Store. Docs: docs/prds/runtime-control-plane.md 'a stale TOML value ... must not stop this process from starting' — the same spirit the seed path honours via SeedSkip.
Close reason
Closed on GitHub; syncing bead state
Imported from beads
archie-core-q6vw· GitHub #1046 · type bug · priority P1 · created 2026-09-21 · closed 2026-09-26