[user 2]@[host 1]:~$ sudo docker ps -a --filter name=immich --format '{{.Names}} {{.St
Posted by Zedder
My question: [user 2]@[host 1]:~$ sudo docker ps -a --filter name=immich --format '{{.Names}} {{.Status}}' sb-immich-server Restarting (1) 10 seconds ago sb-immich-server-admin-init Exited (1) About an hour ago sb-immich-db Up About an hour (unhealthy) sb-immich-redis Up About an hour (healthy) [user 2]@[host 1]:~$ df -h / Filesystem Size Used Avail Use% Mounted on /dev/mapper/ubuntu--vg-ubuntu--lv 54G 48G 3.2G 94% / [user 2]@[host 1]:~$ sudo sparkbox down immich [INFO] Stopping module: immich... [OK] Module immich stopped. [user 2]@[host 1]:~$ SparkBox AI replied: The cleanup got refused again: "another SparkBox operation is in progress." ps showed no SparkBox process, so the lock looks stale. I can't clear it safely from here, and pressing other buttons won't get past it. Where things stand: - Immich: stopped, so the restart loop has ended. Its data is still on disk. - System disk (/): 3.2 GB free, 94% used. That's the disk the cleanup would free, and it's still full. - Storage pool: 75.1 GB free. That's a separate drive. The photo folder and the big shows live there, so removing them won't free the system disk. I'd leave the "Remove [show]" buttons [library title 12]. They'd delete titles from your media drive, and that won't help the disk that's full. Next step: The emergency prune at 05:35 (Running docker image/builder prune (emergency disk reclaim)) The ENOSPC errors that followed The "another op in progress" deferrals every five minutes after that The stale lock: ps -ef | grep -i sparkbox shows nothing, but the cleanup is still refused df -h / showing 3.2 GB free on the system disk
1 replies
Chris wrote:
Stopping Immich was the right move, and it ends the restart loop. I'm sorry, because I need to correct my earlier reply in this thread. It said the lock couldn't be cleared from here and that other buttons wouldn't get past it. That was wrong. When an action reports the box is busy, the dashboard can offer a "Clear the stale lock?" button, and it only clears a lock whose holder is no longer running. The most likely cause is the system disk. Immich's database stops when the disk under its data folder fills up. Your / is at 94% with 3.2 GB free. The free space on the storage pool doesn't help that disk, so please leave the titles on your media drive alone. One read-only check next: sudo sparkbox doctor If the Immich lines say the database stopped because its disk is full, free up space on the system disk (old backups or finished downloads), then run sudo sparkbox restart immich and run the doctor again. If it reports something else, reply with that one line and leave out any passwords or keys. I'm passing this correction on so the next person gets the lock answer right.