gsadmin/pulse
1
0
Fork 0
mirror of https://github.com/rcourtman/Pulse.git synced 2026-10-23 20:25:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(recovery): migrate legacy point schema before indexing

Browse Source
This commit is contained in:
rcourtman
2026-03-27 09:21:28 +00:00
parent a31227620d
commit c49b599739
4 changed files with 675 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/release-control/v6/internal/subsystems/storage-recovery.md
+6
View File
@@ -223,6 +223,12 @@ those shared store reads, including recovery-backed reporting, alert rollups,
and tenant-scoped AI recovery-point adapters. A malformed metadata blob may
degrade row-local enrichment, but it must not take down adjacent readers that
consume the same canonical recovery store.
That same recovery-store migration boundary must keep legacy schema upgrades in
dependency order. When a persisted `recovery.db` predates columns such as
`item_type`, the store must add the migrated columns before creating indexes or
running query paths that reference them, so opening a legacy store backfills
cleanly instead of returning `500` from `/api/recovery/points` or
`/api/recovery/rollups` during schema initialization.
That same hook-boundary normalization also owns the runtime recovery display
model. Canonical recovery points and rollups must expose `display.itemLabel`
and `display.itemType` to recovery consumers, while legacy transport fields