Restore faulty Immich from backup
I have a Immich instance installed on my home server using docker (on Ubuntu server, on an old HDD).
Last month I was uploading several pictures on my Immich, and it was very busy extracting metadata and running AI on the new assets (it was like 1000 new assets). Suddenly the power went out.
Once the power was back my server was no longer booting, I was stuck in Initramfs.
I run
fsck /dev/sda1 -y
and it repaired the faulty file system. Then I booted correctly.
Unfortunately almost all docker container (pihole, radar, ecc…) were working fine but Immich.
Immich (immich_server and immich_postgres) containers were stuck in a reboot loop. Stopping them and restarting didn’t work.
Fortunately I had a weekly backup, therefore I rolled back ALL the subfolders in /var/lib/docker/ (overlay2 included), and I also restored my postgres folder (but not my library folder, I left it as it is).
Still it was not working properly, Immich was rebooting and not healthy in docker.
Then I followed the guide at “https://docs.immich.app/administration/backup-and-restore/“, the part about manual restore.
I restored everything from the *.sql.gz backup archive that Immich automatically created. This worked perfectly and now my Immich server seems to be working!
But I am quite uneasy on this installation now. Is there something on Immich that I should run? Is there something I should do or check? What about the photos that were uploaded after my backup, they should still be in the upload_location but not indexed by the database, is this an issue?
Last month I was uploading several pictures on my Immich, and it was very busy extracting metadata and running AI on the new assets (it was like 1000 new assets). Suddenly the power went out.
Once the power was back my server was no longer booting, I was stuck in Initramfs.
I run
fsck /dev/sda1 -y
and it repaired the faulty file system. Then I booted correctly.
Unfortunately almost all docker container (pihole, radar, ecc…) were working fine but Immich.
Immich (immich_server and immich_postgres) containers were stuck in a reboot loop. Stopping them and restarting didn’t work.
Fortunately I had a weekly backup, therefore I rolled back ALL the subfolders in /var/lib/docker/ (overlay2 included), and I also restored my postgres folder (but not my library folder, I left it as it is).
Still it was not working properly, Immich was rebooting and not healthy in docker.
Then I followed the guide at “https://docs.immich.app/administration/backup-and-restore/“, the part about manual restore.
I restored everything from the *.sql.gz backup archive that Immich automatically created. This worked perfectly and now my Immich server seems to be working!
But I am quite uneasy on this installation now. Is there something on Immich that I should run? Is there something I should do or check? What about the photos that were uploaded after my backup, they should still be in the upload_location but not indexed by the database, is this an issue?
