Unable to run immich after latest upgrade (1.58.0)
After upgrade to 1.58.0, two of the containers can't run and are constantly complaining of these two errors
microservices:
29 Replies
server:
other containers seems to run normally
Just confirming, did you make the required changes to docker-compose.yml?
yes
two changes and one deletion
actually three deletions
but you know what i mean
Okay, not sure if it's exactly the same errors but those two containers kept restarting for me until I brought the entire compose stack down and back up again (I posted in general-discussion)
i have this setup via portainer
does that mean i need to stop the stack, or delete it or what, and then create it ?
I believe just stopping the stack then restarting it, though I'm not 100% sure.
yeah, that didn't work ...
will see what else i can do here
You might have to do compose down then up again
It's worded differently in Portainer though, trying to remember the equiv. of compose down & up
I think it's just the "Stop this stack" then "Start this stack" button within the immich stack, but if that's not working maybe something else is going on for @suntribe
yeah, i'm checking if there's some other wording in portainer
i think you might be right
stop this stack aka down
and start this stack aka up
but still no dice, eh?
no
you were right, start/stop is down/up
but that didn't help
crap, you may have run into something yuckier then. I yield to the immich experts.
tnx for your help
i just rolled back the changes
from entrypoint to command
i rolled back to
entrypoint: ["/bin/sh", "./start-server.sh"]
and in immich-web i left the requested change (deleted the entrypoint)
and everything came back to lifemy server is starting fine, but can not connect to database
[Nest] 1 - 05/27/2023, 11:06:06 PM ERROR [TypeOrmModule] Unable to connect to the database. Retrying (1)...
Error: getaddrinfo EAI_AGAIN immich_postgres
Did not change anything on this
what do logs from postgres container say?
ready to accept connections 🙂
is postgres part of the same network as the rest of the stack?
I have portainer as well. I deleted all the containers, updated the entry point config, the deployed the stack and it all is working fine
don't know what to say, i also did the same, but it complained
You could try restarting immich server container. Maybe the database came up after the server
Did you change entry point to command?
then, i actually went back from
command to entrypoint and that's what brought back the app to life
why? will this cause any issues in the future - i think it might... but i don't know what else to try atmIt's not a big deal either way.
We default the entry point in the docker file now so this is just a change to make it still use the same command to start the container.
i was thinking, maybe there might be some info, some clue, when the 'upgrade` popup appears, that would indicate that a breaking change is in current release and that one should not just blindly update the stack...
Yeah docker versioning is not standardized so I'm not sure how they'd even know if it was breaking change or not. Best to always look over the release notes.
especially, not without updating the docker-compose.yaml file
yeah, i know
Restarted docker service on host helped. Maybe something was broken
Needed to do another pull to get the latest images.