23 days ago
My service kazi_yangu in project ideal-tranquility (Project ID: 63c56d2c-90d6-41c4-bdb8-198b3a4e6a71) had its container restarted after an environment variable change today (May 21, 2026 around 10:00-11:00 UTC). The service had a SQLite database at /app/prisma/dev.db inside the container with 256 production records that was lost when the container restarted. The previous deployment ID is 781c211c-c898-4299-9aee-0458192f885f. Please can you check if the old container's writable layer still exists on your infrastructure and recover the file /app/prisma/dev.db from it. This is urgent production data.
3 Replies
23 days ago
Data stored in the container's filesystem (outside a mounted volume) is ephemeral and does not survive restarts or redeployments. We do not retain old container writable layers and cannot recover files from previous deployments. To persist data like your SQLite database across restarts, you would need to attach a volume mounted at the path where your database file lives.
Status changed to Awaiting User Response Railway • 23 days ago
Railway
Data stored in the container's filesystem (outside a mounted volume) is ephemeral and does not survive restarts or redeployments. We do not retain old container writable layers and cannot recover files from previous deployments. To persist data like your SQLite database across restarts, you would need to attach a [volume](https://docs.railway.com/volumes) mounted at the path where your database file lives.
23 days ago
What is the best path foward
Status changed to Awaiting Railway Response Railway • 23 days ago
Status changed to Open Railway • 23 days ago
hoseajonas
What is the best path foward
23 days ago
You should attach a volume to your service (right click your canvas, add a volume, select your service, and mount it to /data). You should then update your codebase to access the SQLite database from the /data path (eg, /data/database.db).