a month ago
A PostgreSQL 18 service on a Hobby plan has filled its existing 5 GB volume and now crashes during automatic recovery.
The generic error is: FATAL: could not write to file "pg_wal/xlogtemp": No space left on device. Connections are rejected because consistent recovery state has not yet been reached. The dashboard console says "No running instances" and the volume page shows 100% capacity. There is no Railway volume backup listed.
Before the crash, a bounded cleanup exported selected old telemetry to verified off-platform ZIP archives, then deleted only the exact unchanged archived row versions. This did not produce a measured filesystem reduction. Further cleanup and export attempts have stopped because the database is unavailable. We have partial logical archives, not a complete restorable database backup.
What is the supported non-destructive recovery path within the existing 5 GB limit? Can Railway assist with read-only filesystem inspection and identify genuinely disposable temporary/non-database files to create startup headroom, or provide another supported way to bring the database up long enough to resume verified bounded cleanup?
Constraints: no plan upgrade, extra paid storage/services, volume wipe, database reset, manual removal of relation/WAL files, or pg_resetwal. Please preserve existing data. We are asking for recovery guidance, not authorising destructive changes or charges.
This is intentionally a general public request: project/service identifiers, credentials, private logs and application data have been omitted. If Railway staff need account-specific context, please advise an appropriate private route.
1 Reply
a month ago
The Postgres volume attached to your service is already at the largest size Hobby allows (5 GB), and the logs show recovery replaying successfully but then failing to write the new write-ahead log it needs to finish. Within your constraints (no larger volume, no deleting database files, no reset), there is no supported way to give it that extra room. We also don't do manual filesystem inspection or file cleanup on customer volumes. The supported way out is a volume larger than 5 GB, which requires Pro, and volume backups are also a Pro feature.
Your existing data is still on the volume. Nothing we can see shows it has been lost.
Separately, your project also holds an earlier, detached Postgres volume that is nearly full and scheduled for permanent deletion at 21:36 UTC today (2026-10-19). Until then its contents are intact, and you can cancel the deletion with the restoration link emailed to your workspace admins or from the project in the dashboard. After that time it can't be recovered.
Status changed to Awaiting User Response Railway • about 1 month ago
24 days ago
This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!
Status changed to Solved Railway • 24 days ago