Postgres logs an error every 30 seconds for the old DeadAir table #196
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Every
ipxcommand, the healthcheck'sipx statusincluded, runscreate_missingin src/db.rs, which still tries the one-time clean-up of the DeadAir copies from before #181:SELECT path FROM deadair WHERE path IS NOT NULL, treating failure as "already done". The table is gone in production, so the query fails each time and Postgres logsERROR: relation "deadair" does not existwith the statement: 120 in the last hour (docker logs --since 1h postgres | grep -c 'relation "deadair"'), about 2,900 a day, in a log other apps share. Found by the ponytail audit of [Unreleased] on 2026-10-11 and confirmed in the prod check of the same day. Production has run the clean-up (no copies left), and it is the only database, so the block and its testdeadair_drops_its_old_copies_oncecan go.Fixed in
12fadbd: the clean-up of DeadAir's old copies and its test are gone. Deployed at 20:13 UTC; the lastrelation "deadair" does not existin Postgres's log was at 20:13:04, before the new daemon started, and none since.