HTTP 500 Error After Container Restart – Missing isVisible Column in assets Table

Hello, I'm encountering a critical issue after restarting my Immich deployment in a Kubernetes cluster. Here's a detailed description of the problem:

Environment: Immich deployed on Kubernetes, split by immich-server, immich-database, immich-ml and immich-redis pods.

Version of Immich (after problem occurred)
# immich -V
2.2.53


Problem: After restarting the Immich container, the web interface loads but many operations fail with HTTP 500 errors.

Logs: The server logs repeatedly show the following error:
ERROR [Api:ErrorInterceptor~m4hlmtjq] Unknown error: PostgresError: column assets.isVisible does not exist


Investigation
  • I connected to the associated PostgreSQL container.
  • Checked the assets table and confirmed that the isVisible column does not exist.
Database Restore Attempt
  • I have a backup of the database (a
    .sql
    file) taken when Immich was functioning correctly.
  • The SQL file contains the isVisible column in the assets table definition.
    COPY public.assets (id, "deviceAssetId", "ownerId", "deviceId", type, "originalPath", "fileCreatedAt", "fileModifiedAt", "isFavorite", duration, "encodedVideoPath", checksum, "isVisible", "livePhotoVideoId", "updatedAt", "createdAt", "isArchived", "originalFileName", "sidecarPath", thumbhash, "isOffline", "libraryId", "isExternal", "deletedAt", "localDateTime", "stackId", "duplicateId", status, "updateId") FROM stdin;
  • I executed the SQL file to restore the database, but it seems the schema was not correctly updated—the isVisible column is still missing after the restoration.
Request for Help:
Is there a known migration path that should add this column?
Is there a recommended way to restore the database schema when deploying on Kubernetes?
Could this be related to a missing or failed migration step after an upgrade?

Any guidance would be greatly appreciated. Thank you!
Was this page helpful?