Updated Immich to 1.111.0, but still running 1.108.0
Hello everyone,
Following the discussion here : https://discord.com/channels/976934649643294750/1268875108970266644, I have updated my Immich instance which was running 1.108.0 to 1.111.0.
I don't have a high performance machine so it took a lot of time to upgrade.
Tipi is displaying that the version in use is 1.111.0, but when I connect to Immich, it is still running 1.108.0.
I'm not really sure what to do, maybe I can just wait for a new version to come (e.g 1.112.0) and try to update at that time. But if I want to update manually to 1.111.0 right now, I don't think I have other choice than deleting and reinstalling the app and losing my photos.
Unless there is an option to reinstall the latest version of Immich within the GUI of tipi or the CLI ? I couldn't find such an option in the documentation.
Thank you !
Here is my debug output, all looks good actually :
29 Replies
Noo don't delete your photos lol
You could try restarting the app
If it doesn't work there is another option:
You can go to the apps/immich folder in the runtipi directory
Then edit the config.json file
And decrease
tipi_version
by 1
Then the update button should appear
And try to update again
Make sure to backup first before doing anything thoughThanks, that's a good idea.
Before attempting your solution, I tried to perform an in-place update first of immich. It turns out that the command :
./runtipi-cli app update immich
is doing something ! I didn't see it documented, for me I only could see the stop/start options in the doc.
Unfortunately it seems to have failed :
✗ Failed to update app.
Error: Error sending request: reqwest::Error { kind: Request, url: Url { scheme: "http", cannot_be_a_base: false, username: "", password: None, host: Some(Ipv4(x.x.x.x)), port: None, path: "/worker-api/apps/immich/update", query: None, fragment: None }, source: TimedOut }
"TimedOut", not sure what is the issue is here. I don't think it's a network timeout to reach the internet, but maybe since my VPS is super cheap and on its knees all the time, there might be an hardcoded delay for waiting of the update to be completed ?
So I started with your suggestion @Stavros , I decreased the app back to 1.108.0 and restarted it.
So far I have not yet seen the update badge, maybe I need to wait a couple of hours. I restarted already runtipi completely but no go yet.
Whoa, you exposed an IP there! That's seems like a public ip! If it is please edit the message and remove it!
You're right, forgot to remove it 🙂
In any way the cli error seems like an issue on our end
Did your decrease the tipi version?
Ah no you are right, only the immich version. tipi version was 99, I decreased to 98, restarting now Tipi
Did it work?
Unfortunately not, even by lowering the "tipi" version and restarting Tipi then the whole server (linux debian 12), it didn't prompt me to update 😦
Oh shit
I forgot we store the version in the database
Well when you tried to update
Tipi should have created a backup automatically
Can you check in the backups tab?
Unfortunately I had to delete the backups when I first tried the update 😦 There were about 4-5 of them. Maybe I should try to decrement the version one by one and see which one triggers the update ?
Oh no lol
Well......
Revert any changes you made on the apps/immich folder
(if you don't remember just copy the immich folder from repos/reallylongstring/apps/immich)
Then start the app
*do all this with the app stopped
This is what I did, restarted both tipi and immich but still no app update showing I'm afraid 😦
Are you sure immich is not on the latest version
Try to clear cache/open immich in an incognito session
Let me try
Mmh no, in incognito mode with other browser it warns me there is an update :
That's weird
What's the version that the docker compose uses in apps/immich/docker-compose.yml
I see this :
cat apps/immich/docker-compose.yml
version: "3.7"
services:
immich:
container_name: immich
image: ghcr.io/immich-app/immich-server:v1.111.0
volumes:
- ${ROOT_FOLDER_HOST}/media/data/images/immich:/usr/src/app/upload
Alright
Immich bug?
If you stop immich in the tipi ui
And start it again
It should show the latest version
If it doesn't immich bug I guess
Because it seems like it's running the latest version
Maybe. I will try to install Immich on an other instance where I have tipi - will be probably later today.
I’m wondering if other people faced this issue as well 🙂
Fwiw I updated immich to the latest version w/o a problem, so nothing general
@frankynov you do not have a costum config for the immich app by chance?
try to docker pull the image manually
So you have it locally
No it’s a « vanilla » tipi installation. But I think something went wrong in the first attempt of upgrade
Would you have a command for this please ? ☺️ I have to admit docker management is still obscure for me
docker pull ghcr.io/immich-app/immich-server:v1.111.0
And this in the runtipi main folder ? Or I have to cd into Immich folder first ?
makes no difference, images will be stored in a central location anyways
Good to know, thanks ! 🙂
I ran the command, restarted Immich and still running v1.108.0
I really then suspect something is wrong on the machine itself; I also saw some weird behaviour with the FireflyIII data importer for example on that machine.
If we are in a dead end I will reinstall everything from scratch (it's my "test" instance anyway).
On my more powerfull VPS (Oracle one with 4 ARM cores & 23Gb RAM), the fresh install of Immich went flawlessly and is running the latest version
Ok so I did it the hard way, as I was anyway having a lot of issues with that cheap server. I switched to another provider with much more power and more "serious". I could install Immich without any issue.
A bit sad that I wasn't able to figure out exactly what was the issue, but there were probably too many unknowns here.
That being said, I learned a lot about docker thanks to this 🙂