Production Postgres database only shows _prisma_migrations
arveniq
HOBBYOP

a month ago

Hello Railway support.

We have a production PostgreSQL service that suddenly appears to have lost all application tables.

Project: humorous-nature

Environment: production

Postgres service: Postgres

Backend service: crm-backend

Public DB host: yamabiko.proxy.rlwy.net:53836

Internal DB host: postgres.railway.internal:5432

Database: railway

The database currently only shows the table _prisma_migrations.

Before the incident, it had production tables like:

  • User
  • Tenant
  • Lead
  • Conversation
  • Message
  • PipelineStage
  • AiAgentConfig

The Backups tab says there are no backups available.

Can you please check urgently:

  1. Whether the Postgres volume was recreated, reset, detached, replaced, or corrupted.
  2. Whether there is any internal snapshot, volume recovery, or point-in-time restore available.
  3. Whether the current postgres-volume is the original volume for this service.
  4. Whether Railway has any internal recovery option even though dashboard backups are not available.

This is our production database and we need to know if recovery is possible before rebuilding the schema.

The issue was noticed on June 8, 2026, after backend deployments around 18:00 UTC.

Solved

1 Replies

Railway
BOT

a month ago

Your Postgres service does have a persistent volume attached, and it is in a healthy READY state using about 244 MB. The volume was not detached, recreated, or replaced. The Postgres logs from around 18:57 UTC onward show continuous "relation does not exist" errors for your application tables (Campaign, Message, Conversation, etc.), confirming the tables themselves were dropped while the database engine and volume remained intact. The fact that only the migrations tracking table survived is consistent with a destructive migration or schema reset having run against this database, likely triggered by the backend deployment at around 18:08 UTC. Our databases are unmanaged at the DB layer, so automatic snapshots, WAL archiving, and point-in-time recovery are not available to restore your data unless you had configured volume backups yourself. Since no backups were configured, we're not able to recover the dropped tables. Going forward, you can enable volume backups on your Postgres service from the service settings page to protect against this in the future.


Status changed to Awaiting User Response Railway 29 days ago


Railway
BOT

22 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 22 days ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...