Mattaton

Interesting. I recently asked Alex about that and he said we should still be on 0.2.0 even though 0.4.0 is available. I'm using PG14, though. Maybe there's an issue with 16 only?

Any chance you have a backup of the database? You can try to spin up a new postgreSQL container and see if it starts without issue. If so, restore the SQL backup to that container and point Immich to it. If no backup, then we'll have to hope someone more knowledgeable than me knows what that error indicates and how to solve it. I can say I just updated to 6.12.14 a couple days ago and everything still ran afterward with no issue. So it's not an inherent issue with 6.12.14 and postgreSQL. Good luck with it! (I'd also suggest posting in the support channel of the Immich Discord server. Might have better luck than here.)

Mine has done it again 3 times in the past month. And even though last time it was supposedly my use of macvlan on docker, this time I am set to ipvlan. So, it's not that. What you describe with the USB is EXACTLY what happens to me. I have two USB ports on the front of the case and I just bounce between them. Very annoying. Hopefully we can get some help. But yeah I do wonder if it's the USB failing.

Instead of a typical postgresql repo, you'll need to use the repo: tensorchord/pgvecto-rs:pg14-v0.2.0 Mine is set to postgres 14 (which is the officially supported version), but from what I understand other versions will work. Even though 0.4.0 of tensorchord/pgvecto.rs is available, stay on 0.2.0.

The imagegenius immich container (which I believe the SI1 template is based on) has most all of these breaking changes covered for you. It removes some of the flexibility/configurability of the official immich container(s) in exchange for ease-of-maintenance. Which is a-okay with me. Been rocking it since day one. 🙂

Back when I first set up SWAG, I looked at NPM, but ultimately went with SWAG. I can see how it might be a bit friendlier for webhosting. But you're still saying spin up a separate Apache for more complex content? If I'm still doing that I'll likely just continue using SWAG for my reverse proxy and use it to point to sites hosted on whatever web serving solution I end up with (still trying to figure out exactly what that is 😄 ).

Initially I was looking for a web server container, but I couldn't seem to find one and ran across a reddit thread talking about SWAG already being a web server, so use that. I like the idea of containers having a specific purpose (kinda the point of them). Any suggestions for an Apache/PHP/MySQL container? Or should I run the DB as a separate container? I'm not super well-versed in web servers. Mostly only used XAMPP for development and then posted to linux-based web servers with control panels like cPanel. The nitty gritty of setting one up as individual pieces is pushing my knowledge a bit. 🙂

I've been using SWAG for reverse proxy for quite some time, works great. I get the idea that SWAG (nginx) is basically a web server with php. I have a local php website I'd like to run off of it, if possible. It's currently running on XAMPP for Windows. Anyone know of any guides to using SWAG as a web server beyond a reverse proxy? Thanks!

Sorry if this has been answered before - I've tried searching the thread, but I can't find anything or I'm not searching for the right thing. I want to run the VM with a graphics card, but without a monitor (I'll use NoMachine to connect). Changing from VNC obviously removes the VNC console, so I then can't select the boot drive. How can I have the proper drive autoselected after say 20 seconds? (I don't want it to instantly select the macos drive just in case I ever need to do something different on that screen). EDIT: I guess my problems are deeper than I thought. If I enable the graphics card and a PCIe USB card in the VM, only the graphics card works. If I just enable the USB, then an attached keyboard works and I can press Enter to boot up. If I plug a monitor into the graphics card, the graphics work, but then the USB doesn't and I can't press Enter to boot. Ugh.

What happens if you try to go here: http://192.168.1.249:8080/vnc.html?resize=remote&host=192.168.1.249&port=8080&autoconnect=1

I think in order to have the webui button, you have to use the template in unRAID (dockerman), enable advanbced view and then fill in the webui field. Looks like you are doing everything via terminal instead. Is there some reason you're doing that instead of using CA?

I just added volumes to my template and those are visible in the root / view. Not quite as fast as having it open in that directory, but just one click away. 🙂

Totally agree that confusion is bad! You're right, setting the TZ env manually is redundant. I took it out of the template. 🙂 I don't know. And I mean that, I truly don't know. All I do know is I've had a world of trouble with timezones in LB. I feel like it's working now and it's like a house of cards that I don't want to breathe too hard on because it might stop working again. I won't rehash it all again since you have heard it before. I know it SHOULD work with just the TZ that unRAID puts in, but it didn't and still has CET sprinkled here and there. I can check the timezone of the container, and it's correct. But still, the logs and interface both have CET offsets. I'm just glad my cron runs at the correct time. I can deal with logs being wrong. 🙂

Gotcha. You're right. I end up seeing it twice in the docker run. I was just pointing out the only two places I have "set" a timezone for LB. But yeah, setting the TZ env is not necessary in unRAID.

The timezone in LB has issues. When I first tried using it 6 months or so ago, I had to set the time for the backup to compensate for the timezone offset. For me, that's 6 hours (EST vs CET). So, I'd set it for 11pm to get it to run at 6am. Everything in LB really wants to force or translate to CET. So you might want to test to see if the cron works if you compensate for your offset. Now, after ditching it and setting it up again a month ago, I can now get it to run at the proper time for me without compensating, but the logs will still say the CET equivalent. So, I have it set to run at 6am EST, it'll say it ran at 12pm CET in the logs, but on the interface it'll say 12pm EST - so it mixes the hour of CET, but still marks it as EST. But, it DOES run when I want it to - matching the time in the cron setting. I don't even bother looking in LuckyBackup interface for the log. I just check the log files. So, how my setup probably differs from yours is that back the first time I was setting it up, I ended up adding a .profile file in the LB config folder which just contains this: export TZ=America/New_York Also, make sure to set the TZ env variable in your unRAID template. LuckyBackup is very nice, but yeah, the timezone stuff gets a bit fussy.

It kinda seemed like it was the same plugin, especially how CA pushed Folder View for folks upgrading unRAID beyond the version Docker Folder supported (functioned on).

Their video was on technically a different plugin. Once Docker Folders went out of development, Folder View more or less replaced it, but they're not the same plug-in.

I've had this problem with postgresql (and other containers) for a while. Are you seeing that the owner of postgresql config folder also changes to UNKNOWN? I have yet to understand why this happens. I have set up a user script (user scripts plugin) that changes the owner to nobody each night before backups are set to run.

I have this issue on several containers. Since there's no way in the container to explicitly set the UID and GID, it just goes to UNKNOWN being the owner? I have set up userscripts in the past that just reset the ownership every day. Very annoying and I don't know if this is expected behavior or not. 😕

UPDATE: Moved the USB from one of the USB ports directly attached to the mobo to a port on the front of the case and that allowed it to boot up. I don't really like connecting the USB on the front like this. But I guess as long as it works. Dockers, etc. started back up after it booted. So I THINK I'm okay. I hope.

Through a perfect storm of circumstances, things have gone from bad to worse, to worse still. Started with uploading a ton of photos to immich container. Didn't pay attention to the fact that I was doing it at a time that would coincide with a parity check. That, in turn, stopped mover from being able to do its thing, so then the cache filled up. That caused all sorts of issues. So I waited for the parity check to finish, then ran mover, and finally attempted to stop the array to see if I could get docker to start back up again (failed to start docker service - assumed caused by full cache). Array would not stop with the trying to stop disk shares issue. So I then powered the system down via the power button in the GUI. Now the system won't boot. I get the following readout on the screen. I have tried reseating the USB and putting it in a different USB port. Same result. Any help is greatly appreciated. Stressed out!

The immich discord server, not unraid. 😄

If you haven't already, this would be a great request to bring to the Discord server. There's a support channel there that would be best equipped to help you with something like this.

I don't have those vars at all on my set up. Not just blank, but removed (never had them). I believe there are built in default for them, so you only need them if you need to change the setting. Which it sounds like you don't. I've had to change the template once, maybe twice, in the time I've had it running. Generally very low maintenance. Even with some of the breaking changes in immich, they've kinda done the work for us with the AIO. I have updated the container every time there's an update with no issues at all. Trying to recall, but I think the biggest changes I've had to make were switching to the vector postgresql repo and changing to the cuda tag on the immich repo (because I have an Nvidia GPU). Best advice would be to check the release notes each time there's an update to make sure there're no hoops to jump through. So far that's told me everything I need to know to keep it spinning like a top. 😊 Oh, and definitely join the Discord server. Very helpful folks over there. Not that there aren't helpful folks here, but the Discord server is very active with immich as the main focus.

I don't have these vars on mine. I set it up long before SI1 did his video. I THINK maybe that's only for setting up an external ML container outside of the AIO imagegenius container. I could be wrong, though. All I know is mine works. 😄 Did a quick search on the Discord channel and I saw this post from June. Apparently those vars have been changed/deprecated? * for ML I did just realize I was still using MACHINE_LEARNING_PORT and MACHINE_LEARNING_HOST as variables and the breaking changes say it's IMMICH_PORT and IMMICH_HOST: 2. Environment variables SERVER_PORT, MICROSERVICES_PORT, and MACHINE_LEARNING_PORT were renamed to IMMICH_PORT HOST and MACHINE_LEARNING_HOST were renamed to IMMICH_HOST " And a reply from mertalev: " You're not supposed to change these envs unless you have a crazy setup Remove these envs and reset ml settings in the Immich admin section if you have changed something here " So I think you can leave those out.