Postgres crash loop - volume resize not propagating to filesystem
Anonymous
HOBBYOP

2 months ago

Hi, my Postgres instance is stuck in a crash loop and I need urgent help.

What happened: My volume hit 92% capacity. I ran a DELETE FROM on a large table which generated WAL logs that filled the remaining space. Now Postgres cannot complete startup recovery because it can't write to pg_wal.

Error: FATAL: could not write to file "pg_wal/xlogtemp.26": No space left on device

What I tried:

  • Upgraded volume from 1GB to 5GB (2+ hours ago) but filesystem still shows no space
  • Added start command to delete xlogtemp files but Postgres creates a new one during recovery and hits the same error

The volume resize does not seem to have propagated to the filesystem. I need Railway to expand the filesystem on the volume so Postgres can complete WAL recovery and start.

Project: glorious-dedication

Service: Postgres

Volume: postgres-volume

Closed

1 Replies

2 months ago

Please do not open duplicate threads.


Status changed to Closed brody about 2 months ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...