Yes, the local filesystem inside a Railway deployment is ephemeral, so SQLite files, JSON files, and any other data written to it are discarded on every redeploy. None of your services currently have a persistent volume attached, which is exactly why your trading history disappears. To persist data across deployments, add a volume to your service and configure your application to read/write from the volume's mount path, or use an external database like PostgreSQL.

Hello,

Thank you for the explanation.

This is very helpful.

Before I modify my application, I would like to confirm one thing.

If I attach a Persistent Volume to my Railway service and configure my application to store its SQLite database and JSON files inside the mounted volume, should all data survive future deployments and restarts?

Is there anything else that must be configured besides:

attaching the Persistent Volume,

writing all persistent data to the mounted volume path?

I want to make sure I follow Railway's recommended architecture for long-term persistent application data.

Thank you.

That's correct. Once a volume is attached and your application writes to the volume's mount path, data persists across deployments and restarts. The only thing to ensure beyond those two steps is that the mount path configured in Railway matches the path your application actually writes to, as a mismatch would cause writes to land on the ephemeral filesystem instead. No other special configuration is needed.