Cloudron Forum

I'd like to restore the immich database to a non-cloudron server. The Immich backup/restore instructions
say to do this:

gunzip < "/path/to/backup/dump.sql.gz"
| sed "s/SELECT pg_catalog.set_config('search_path', '', false);/SELECT pg_catalog.set_config('search_path', 'public, pg_catalog', true);/g"
| docker exec -i immich_postgres psql --username=postgres # Restore Backup
docker compose up -d # Start remainder of Immich apps

Can I use Cloudron's "postgresqldump "backup to do this? It does not appear to be a gzip file. "gzip: stdin: not in gzip format".

Or is there an easier way to restore the database on a non-cloudron server?

The Cloudron docs say that the settings are coming from /app/data/immich.json; however, the file only contains {}. Where are the OAuth settings coming from? I don't want to copy the Immich defaults into the settings file, which presumably would break OAuth.
...
I was able to export the settings .json from the UI, and paste them into /app/data/immich.json, and then edit there.

@girish, Another use case is for those, like me, who spend days on a fully locked down and logged office computer. It would be nice to be able to occasionally circumvent that for browsing without tracking fear.

I've updated to 7.7 and also updated Immich. I can log in to the Immich web page with OAuth, but in the mobile app, it just spins with the log details:

Null check operator used on null value.

Thanks for any help

Thanks, could not have been easier.

I thought I was good after the recent upgrades, but am now not seeing pics in the web interface and seeing in the log:

ERROR [ExceptionsHandler] pgvecto.rs: pgvecto.rs must be loaded via shared_preload_libraries.

Though the android app still appears fine.

Was there anything that we needed to do with pgvectors.rs?

Thanks

I have a CIFS mounted, but it has fixed itself and appears to work normally now.

My Disk Usage panel only shows a spinner. I think this started after updating to v7.6.3 today.

it's been a month, time again to update?

Hi,

Metadata (m:) searches work fine, but standard searches return 500 - Internal Server Error.
The machine learning container location may need to be updated in the compose file?
That's my best guess from looking at similar issues on Discord.
I am also seeing Matplotlib permission errors in the logs.