18 days ago
My PostgreSQL database is stuck in a crash loop with: FATAL: could not write to file "pg_wal/xlogtemp.30": No space left on device
I've already resized the volume from 500MB to 5GB via the UI and tried multiple restarts and redeploys, but Postgres still can't complete WAL recovery. I also attempted a backup restore but Postgres keeps crashing.
Project ID: 7ba6ce3d-19ef-4ec9-8fb3-cbcea40cb786
Could someone help grow the volume or clear the WAL files so recovery can complete?
8 Replies
17 days ago
It seems you've mounted a volume created from a backup - can you remount the original volume so we can fix this for you?
Status changed to Awaiting User Response Railway • 17 days ago
17 days ago
I have two volumes: the original 'postgres-volume' (resized to 5GB) and the backup restore volume 'postgres-2026-02-25 06:09 UTC' (500MB). The original is the one that has my data but can't complete WAL recovery. How should I proceed — should I delete the backup volume so the original remounts, or can you handle it from your side?
The one on the left is 500 MB, the one on the right has been resized to 5 GB but not mounted to anything - a bit confused as to which to work with / restore.
Attachments
Status changed to Awaiting Railway Response Railway • 17 days ago
17 days ago
We've escalated this to the platform team to ensure the 5GB resize is applied to your original "postgres-volume". Please remount that original volume to your Postgres service (you can detach the backup restore volume first), and we'll get the filesystem resize sorted on our end.
Status changed to Awaiting User Response Railway • 17 days ago
17 days ago
Hey, so when I click the postgres-2026-02-25 06:09 UTC", 500MB volume section (the one that is infinitely loading saying 'restoring backup' - I dont see an option to detach - all I see is the attachedIt also says that I can't delete the volume because I will need to delete the service first?
If possible, can you send some detailed instructions with screenshots - it's my first time using your guys service and I am not really well versed here.
Or if you are able to detach it from your side and remount the original volume that would be even better.
Thanks again.
Attachments
Status changed to Awaiting Railway Response Railway • 17 days ago
17 days ago
Ignore the above, I think I sorted it, Iremounted the original postgres-volume and Postgres is back online. Can your platform team still apply the filesystem resize to ensure the full 5GB is available? It is still crashing (see attached)
Attachments
16 days ago
To avoid any more back and forth, some more info - after I reattached the original postgres-volume as instructed. It's still crash-looping with 'No space left on device' during WAL recovery. The volume shows 5GB in the dashboard but the filesystem was never expanded. Hopefully the filesystem resize works. This has been down ~24 hours.
If that does not work can you restore from the backup (the 2026-02-25 06:09 UTC snapshot) into a fresh Postgres instance with the 5GB volume pre-provisioned?
Really hoping we can get back online quickly here and avoid another back and forth.
Thanks again for the help.
16 days ago
Hey! Thanks for re-mounting the original volume.
I've fixed the resize issue and re-deployed your Postgres service so you should be up now. Can you confirm that everything works as intended?
Status changed to Awaiting User Response Railway • 16 days ago
16 days ago
It is working now - thanks for the help guys!
Status changed to Awaiting Railway Response Railway • 16 days ago
Status changed to Solved sam-a • 16 days ago