Repository navigation
fix(infra): raise PostgREST mem cap 96m→256m (OOM killed NAV ingest) - #199
Merged
Merged
Conversation
The 2026-06-03 supabase fleet re-up (project rename, PR #190) applied compose.prod.yml's rest: 96m cap for the first time. Every NAV batch upsert (POST /nav_postings?on_conflict=id) since then blew the cgroup limit: the kernel OOM-killed postgrest mid-request (RestartCount=35), Kong returned 502, backfill-nav 500'd, and NAV ingest died silently with data frozen at 2026-06-01. 256m was applied at runtime via docker update on 2026-07-10 and verified by draining the full 95k-posting backlog; this commit makes it survive the next fleet re-up. Co-Authored-By: Claude <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Raises the
rest(PostgREST) memory cap in compose.prod.yml from 96m to 256m.Why
The 2026-06-03 supabase fleet re-up (compose project rename, #190/#192) applied the 96m cap to a freshly created
kiba-supabase-rest. From that moment, every NAV batch upsert (POST /nav_postings?on_conflict=id— a 50-page backfill body) exceeded the cgroup limit. The kernel OOM-killed postgrest mid-request (dmesg:Memory cgroup out of memory: Killed process (postgrest), RestartCount=35), Kong returned 502, the daily 06:00backfill-navtick 500'd, and NAV ingest died silently with data frozen at 2026-06-01.This went unnoticed until 2026-07-10 because the root disk then also filled (separate issue), taking Postgres down entirely and blanking the landing page.
Verification
docker update --memory 256m kiba-supabase-restapplied at runtime on 2026-07-10backfill-navsucceeded and drained the full ~95k-posting backlog (two ticks,completed:true), snapshots recomputed for 2026-07-10, landing-page Arbeidsmarked/Oppstart cards render again🤖 Generated with Claude Code