#525 ·
agent-opsReview evolutiva backup policy + space (2h/30d/gdrive, ~1.75GB steady-state; context: WI #509)
- Ref
#525(#525)- Project
agent-ops- Status
- backlog
- Priority
- normal
- Type
- todo
- Assigned
- nw-whey-cc
- Created by
- —
- Created
- 2026-05-23T06:32:17.810Z
- Updated
- 2026-06-03T04:06:03.193Z
Questions
No questions.
Event log
-
wi cli
-
Diagnostic (2026-06-03, 10-nightwatch): movilba_dev .dmp dumps in gdrive:remotedocs/backup/movilba_dev/ accumulated unpruned back to Apr26, ballooned ~2.7MB->42MB on May29. ROOT CAUSE of no-prune: evolutiva-backups/run.sh retention (GDRIVE_RETENTION_DAYS=2, find -name '*-PROJECT.dump') only matches *.dump, NOT the movilba *-movilba_dev.dmp pattern -> movilba dumps never pruned. Producer NOT on whey (/var/backups/postgresql empty, no root cron, backup-movilba-db.service points at missing script); likely on lezama dumping dbcaba 10.9.10.54 - confirming with nw-lezama-cc. RESOLVED for now: Elazar emptied Drive trash, account 37GB->3.4GB used, movilba_dev folder gone. Elazar approved (a) retention to stop refill + (b) weekly Drive-usage watchdog email. Impl owner=bin-whey-cc (gdrive-side) or at source if producer on lezama. Also surfaced: whey gdrive-bisync wedged 36h needing --resync (separate fleet /gdrive redesign led by nw-lezama-cc).