▣ wi

Pluto: Supabase free-plan egress fix (backup 30min→8h + data-only ephemeral excludes) + DB hygiene (notifications TTL, cron.job_run_details prune, drop stray bench schema)

Ref

PLUTO-37 (#905)

Project

pluto

Status

done

Priority

normal

Type

task

Assigned

—

Created by

wi-cli-venus

Created

2026-06-11T02:56:31.870Z

Updated

2026-06-11T02:59:08.045Z

Closed

2026-06-11T02:59:08.045Z

Sub-items

Questions

Event log

• 2026-06-11T02:56:31.870Z · created · wi-cli-venus
• 2026-06-11T02:57:07.846Z · note · wi-cli-venus
  From Elazar cross-project brainstorm; root cause (Mars diag) = backup EGRESS, not DB size. Pluto DB 28MB/5.7% — never size-bound. HOST FIX DONE+VALIDATED (nw-whey): evolutiva-backup@pluto 30min→8h (48→3 dumps/day, 16×) + --exclude-table-data notifications/auth.sessions/flow_state/refresh_tokens; dump 1.07MB→0.84MB (−21.6%); run.sh 2.2.2; egress now ~1.6% of prior, under 5GB cap. DB HYGIENE: notifications-read-prune cron LIVE (jobid=5, nightly 03:23, read>30d; 72 rows cleared). appEvents-retention-prune already live from round-1 (jobid=3, tiered 90d info/warn·180d error·365d security) — cited as the evolutiva-commons §DB-Hygiene reference impl. cron.job_run_details 7d-prune RELEASED to db. bench schema (stray synthetic 10k-row artifact, ~1MB) DROP routed to audit pre-apply. Level split 79% info / 17.8% error / 3% warn — info-dominant, conforms, no warn-skew. pluto-commons §Backups doc fix (2h/30d→8h, gdrive 2d/USB 90d) routed to bin-whey-cc. Remaining: cron-prune jobid confirm + bench audit-PASS+apply.
• 2026-06-11T02:58:36.147Z · note · wi-cli-venus
  Docs landed: pluto-commons §Backups corrected live by bin-whey-cc (8h cadence + data-only excludes + gdrive 2d/USB 90d; gitignored pluto/ tree so live-on-disk, no SHA). evolutiva-commons §Free-Tier Backup & Retention Hygiene rule pushed SHA a521db4 (Pluto's appEvents-retention-prune cited as reference impl). cron-job-run-details-prune live (jobid=6, 7d). ONLY remaining item: audit pre-apply review of DROP SCHEMA bench CASCADE → db applies on PASS (~1MB). All else done.
• 2026-06-11T02:59:08.045Z · completed · wi-cli-venus
  DONE — Pluto Supabase free-plan fix complete, all lanes. Root cause = backup egress (not DB size, Pluto 5.7%). HOST: evolutiva-backup@pluto 30min→8h + data-only ephemeral excludes (nw-whey, validated, 16× fewer dumps + −21.6% size → egress ~1.6% of prior, under 5GB cap). DB HYGIENE: notifications-read-prune (jobid=5), cron-job-run-details-prune 7d (jobid=6), DROP SCHEMA bench CASCADE (audit PASS:bench-drop, ~1MB). Pre-existing appEvents-retention-prune (jobid=3, tiered) cited as the evolutiva-commons reference impl. DOCS: pluto-commons §Backups corrected (bin-whey, live); evolutiva-commons §Free-Tier Backup & Retention Hygiene pushed a521db4. No paid tier needed. Cross-project initiative (mars/venus/pluto) all closed.