Failed jobs
Hey guys. I posted this on reddit the other day and was told if it continues to post it here. I have a recently uploaded a library of about 50,000 photos, which were exported from Google Photos.
My job queues show varying amounts of failed items. In particular the 'Generate Thumbnails' and 'Extract Metadata' queues are my main concern. They show 11,389 and 7.549 failed respectively.
As was initially suggested, I re-ran the 'Missing' option on the generate thumbnails queue, and left it to complete for several days. I just checked on it and the queue is now empty (was about 38k a few days ago). The number of failed doesn't seem to have budged though.
I'd already paused all the other queues apart from the Thumb and Meta queues, and since then nothing has appeared in my log files of the microservices docker log.
How do I get it to re-process these failed items?
Appreciate any help/suggestions. I'm happy to gather any other logs or info that might help. Just tell me where to look.
4 Replies
what is the resources that you allocate for Immich?
I also faced with failed jobs badges. After inspecting code and logs I have come to conclusion that failed jobs are left in redis forever.
I would suggest to recreate redis container and rerun jobs for missing items again
It's running on a QNAP TS-230 NAS, Container Station. I think the NAS has a realtek quad core 1.4ghz and 2gb of ram. It only has 2x HDD, and I think it might be maxed out on disk IO a lot, as the CPU and RAM never seem to be an issue.
I think there should be sent a request to the Redis instance to delete the failed queue when a new job is started. In that case you will know that the 'failed' jobs are always a "fresh" number.