controller v0.296.0: a backup run cut off by a power cut or restart is said on the backup pages (R-519); the whole-system backup text states today's measurement (R-518)
gates / gates (push) Successful in 30s

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0159rPz1ZhFKsS53msqPYxtS
This commit is contained in:
2026-10-05 12:14:54 +02:00
parent 477e2548db
commit ff69074514
21 changed files with 2700 additions and 21 deletions
+19
View File
@@ -1,3 +1,22 @@
## v0.296.0 — a backup cut off by a power cut or a restart is said on the page; the whole-system backup text states today's measurement (R-519, R-518) (2026-10-05)
**MinAgent: 0.131.0** (unchanged). New household string: `backups.interrupted_run` (Hungarian and English); the
whole-system backup text and its confirm changed (`backups.a_mentes_alatt_az_alkalmazasok`,
`backups.elinditod_a_teljes_rendszermentest_most`).
- **R-519.** `internal/backup/run_record.go`: the app-data run is on record (`appdata-run.json` in DataDir, atomic, at
both ends — the restore record's shape, R-550). A start that finds a run still marked running turns it into an
INTERRUPTED notice on /backups and /backups/apps, kept until a run ends with every step OK (a run with a failed step
keeps it). Measured 2026-09-14 (BIGNIGHT F2): after a power cut mid-volume-dump both pages said nothing and „Utolsó
adatbázis mentés … OK". After a restart that line is SYNTHESISED from the dump files — now it reads failed when the
newest run was cut. The restore point's own time is already its OLDEST part (data block, v0.275.0, R-696). Wired in
`main()` (`loadRunRecordAtStartup`, AST-pinned). Tests `TestRunRecord_*` (incl. the production path through
`runDBDumpsInternal`), parity cases `backups_interrupted_run`, `backups_apps_interrupted_run`; 3 red-proofs.
- **R-518.** The whole-system „Mentés most" text and confirm now give both measurements — ≈6 minutes on a 9-app box
(demo-hp today: stopped 09:19:08Z, last app back 09:24:55Z, only the local tier ran) and ≈8 minutes on a 12-app box
(2026-09-14) — say "minutes, not seconds", and that the off-site copy in the same run makes it longer. 1 red-proof.
- Evidence: `felhom.eu/documentation/audits/hub-safety-2026-10-05/partE/`.
## v0.295.0 — a box that was off at its backup time catches up once; the household sees why in a banner (R-871, `09` decisions 109–110) (2026-10-05)
**MinAgent: 0.131.0** (unchanged). New household strings: the timeline line `event.backup_catchup_done` and the banner