Railway Postgres: Could Not Locate a Valid Checkpoint Record

How to fix it

Try these in order — the most common fix is first.

1. 1. Restore instead of resetting

   Copy steps
   1. If point-in-time recovery is enabled, restore to a new service
   2. Otherwise restore a volume backup from the Backups tab
   Did this fix it? Yes No

2. 2. Last resort: pg_resetwal

   Copy steps
   1. Create a manual volume backup in the Backups tab
   2. Set Custom Start Command to 'sleep infinity' and deploy
   3. Run railway ssh and confirm PGDATA and matching Postgres major version
   4. Preview: su postgres -c "pg_resetwal --dry-run '$PGDATA'"
   5. Apply: su postgres -c "pg_resetwal --force '$PGDATA'"
   6. Remove sleep infinity, redeploy, then immediately dump the data and restore into a fresh Postgres service
   Did this fix it? Yes No

Symptoms

PostgreSQL fails to start; logs show 'FATAL: could not locate a valid checkpoint record' or 'invalid primary checkpoint record', sometimes at PANIC level.

Why it happens

• Damaged WAL segment or control file, so no valid checkpoint exists to recover from

This guide also covers

• railway postgres could not locate a valid checkpoint record
• railway postgres won't start
• railway database crashed

Other Railway errors

• Railway Application Failed to Respond (502)
• Railway: No Start Command Could Be Found
• Railway Slow Deployments and Slow App
• Railway Node.js SIGTERM Handler Not Running
• Railway ENOTFOUND redis.railway.internal
• Railway SSL Certificate Stuck or Not Working (526, 525, Too Many Redirects)
• Railway 405 Method Not Allowed on POST Requests
• Railway Services Stopped: Trial Credits Ran Out