Weird TCP when trying to load Shared Album
im jusing windows docker-compose using docker desktop any ideas. I get this when loading my reverse proxy and only upon loading shared albums.
60 Replies
i get it on local pc connection as well
Can you help check the server log?
and can you help posting your setup?
tyes sorry. do you want my docker-compose for the setup. it just says 400 error in the docker container. can you specify where i can find the logs?
im using version 1.49.0 on Windows 11 Docker-Desktop
from the command line you can do
docker logs immich_serverhmm everything looks fine
perform docker-compose down
then docker-compose up
from the directory where you store
docker-compose.yml filesame error but a new error
How much RAM do you have on this Immich instance?
@Zack Probably related to https://github.com/immich-app/immich/issues/1868
GitHub
[BUG] Album Page takes way to long to load · Issue #1868 · immich-a...
The bug If I open the Albumlist with currently about 7 Albums it takes about 6 seconds to load each time I open the Album overview or return from an Album. It looks like it takes so long to generat...
Yea, likely. I think that's due to our recent changes in the typeorm entities to return everything all the time. We should now go back and amend the DTOs and the queries so they specifically don't request all relations in certain requests
i have set aside 8gb dedicated to ram
Looks to me that whichever container that is gets to 2GB of ram then falls over
Is the 8GB shared between all containers?
regardless, I will push out a fix next release
kitchenowl barely uses 512 mbs its just a grocery list app. and nextcloud is basically never used but never goes above 2gb
do you know how i can fix it in the meantime
How many assets do you have on the instance?
how can i figure that out?
Go to administration
server stats
in docker or immich?
In the web UI for immich
Hmm not that many, probably lots of album causes the issue
i only have 1
I would say for now that there's not really a huge amount you can do, there's just an issue with the queries we make being not very optimised. For a bit of context, we're going through a huge amount of refactoring in the server to prepare for some big changes, so some of these have introduced some degradation in performance
How many assets in it?
like 1000-2000 images and videos?
That'd do it
i cant see right now because it wont load
You can check from the mobile app I believe
so how can i fix it in the meantime it was working on 1.48.1 with no issues
from the Library tab
no none of the albums load on any apps or web
it will list how many assets are there
I would imagine the mobile app will also crash the server
ah
okay
OK I will aim for a release soon to address this issue
For now, not fixable, and you won't be able to downgrade to 1.48.1 really due to database migrations
Will have to wait for a bugfix to be released
in the mean time, I am not sure why your container would crash as 2GB of ram usage though
no worries. Do you know how many days or weeks. I'll look into it i set a dokcer limit to 8gb for docker because it was eating ram for no reason
i may have limited the docker to 2gb per docker container ill look into it
you have dedicated 8GB, in theory it should be able to push beyond the 2GB unless there is hard setting in the Node process itself
Pretty sure node by default has limits
2GB sounds about right
Days for sure
I might take a look now to be honest 😛
if i fix that limit will it fix most of my error temporarily?
yeah if Zack can push out a fix I can push out a release sometimes tomorrow
If you increase the RAM limit, it's possible it'll manage to survive on the extra RAM 😛
cool thank you for the great work. I love the app and appreciate all your teams hard work
its set at 8GB right now how much should i allocate
For the server specifically it's set to 8GB?
If so, I imagine Node.JS itself is limiting to 2GB
You can fix that by setting and environment variable
NODE_OPTIONS=--max-old-space-size=4096
That should raise the limit to 4GB
it looks like my swap is only set to 2gb but the mem in my ubuntu wsl2 which is how docker is running is set 2g
this says it only takes 50-80% of your max vmem
Limiting Memory Usage in WSL2 | Aleksandr Hovhannisyan
By default, WSL 2 may reserve a significant chunk of your host RAM. Thankfully, we can limit its memory usage with a .wslconfig file.
should i do this in the .env files or the compose?
I would do it on the server container specifically
so how would i go about doing that also if i update the image it will lose that setting correct if its not in the compose?
like so
That should do it
okay one moment let me try
i got in that time thank you!!!!!
any idea on the update status?
yeah it is in, I am planning to push out an update either tonight or tomorrow