I ran a redeploy on a Planka service that used the template in Railway. It was set to latest tag so today it pulled version 2.0.0 into the existing version 1.2.6.

In doing so the Planka service now just crashes after being restarted each time.

The Deploy logs show: Error: The migration directory is corrupt, the following files are missing: (with a large number of named file names listed afterward)

The Postgres service and its data seems to remain as it was.

It looks like some problems with the new data model/files/directories expected in the migration/upgrade to version 2.0.0 from 1.2.6 and the Planka site seems to say this is irreversible.

Is there some way to extract the data in PostGres and import it into a new instance of Planka? I can see data by viewing the Postgres service from Railway, it looks intact but I don't see anyway to extract it. The Planka app won't run without an error. Can I do something with the Railway tools to extract and import it or reconnect the Postgres to a new working instance of Planka, whether it is using the older 1.2.6 version or another?

I have tried reverting to older deploys in Railway, but these all crash after deployment.

Thanks for any pointers.

I don't have a backup, as I found out the Hobby plan does not allow for any backups?

Marc Goblot