Joloxx9

This is onw of this cases when something simple has been lifted to the limits and now it is unusable. Good luck trying to do anything:

It does work now, it looks like GPU group changed and immich could not access it anymore, after adding immich to the new GPU group all is running well.

I just found this post: https://www.reddit.com/r/unRAID/comments/1tbnlfx/730_stableintel_arc_310_hw_transcoding_stopped/ I am on 7.3 RC1, and I was not aware about it until now, this user is correct, ARC card doesn't work, in my case it is A380. It is detected in the system, but Unraid does not read anything from it (intel_gpu_top returns igpu only). I have set it in plex as transcode device and it does fall back to igpu. Immich fails to use GPU as well got errors in the log when it is trying to use it for AI. Device is listed under System Devices. Please see attached Diagnostics. arrakis-diagnostics-20260513-0713.zip

Well in this case, I will have to set up everything in IPMI as nothing besides "coretemp" sensor is visible in unraid, also I did following command: modprobe nct6687 force=1 manual=1 And after this I could finally see some: NCT6687-ISA-0290 A lot of readings but most of them inaccurate like 0, 255 etc but some fan's are visible. Is there anything I could do? This is my motherboard: https://www.asus.com/uk/motherboards-components/motherboards/workstation/pro-ws-w880-ace-se/

Hi mate, are you able to compile a new version please? I have an issue with fans on W880-ACE-SE

Actualy I found the issue, I have share mounted from my other NAS, if I power it down without unmounting first in unraid, then it won't be able to stop the array.

I was trying to stop the array, like always I turned off the docker first, then after few minutes I pressed Stop Array, it stuck like that for 30 min, I was able to log via webgui/ssh etc I just could not stop it. I really need to get to the bottom of it as I do not shut my server ofter(once every 1/3 months) but I need it to be reliable I cannot do parity checks everytime i shut it down because I have to force a shutdown. Please see attached diagnostics - generated when it was stuck, also this was the end of the LOG:</ Apr 20 16:02:43 Arrakis usb_manager: Info: rc.usb_manager Reset Connected Status Apr 20 16:02:43 Arrakis emhttpd: shcmd (4532053): /etc/rc.d/rc.libvirt stop Apr 20 16:02:43 Arrakis rc.libvirt: Stopping libvirt daemon... Apr 20 16:02:43 Arrakis rc.libvirt: virsh net-destroy 13f20c73-b89c-4e11-9a12-68458ae72e02 Apr 20 16:02:43 Arrakis dnsmasq[28282]: exiting on receipt of SIGTERM Apr 20 16:02:45 Arrakis rc.libvirt: libvirt daemon... Stopped. Apr 20 16:02:45 Arrakis rc.libvirt: Stopping virtlog daemon... Apr 20 16:02:46 Arrakis rc.libvirt: virtlog daemon... Stopped. Apr 20 16:02:46 Arrakis rc.libvirt: Stopping virtlock daemon... Apr 20 16:02:47 Arrakis rc.libvirt: virtlock daemon... Stopped. Apr 20 16:02:48 Arrakis emhttpd: shcmd (4532054): umount /etc/libvirt Apr 20 16:02:48 Arrakis kernel: BTRFS info (device loop2): last unmount of filesystem 63a062d1-12e2-4021-b2c9-610bd74f8d93 Apr 20 16:02:48 Arrakis ttm-plugin: Killing all Tmux sessions accessing the array/pools (as configured)... Apr 20 16:03:19 Arrakis unassigned.devices: Remote server '192.168.50.153' port '445' is not open; server appears to be offline. Apr 20 16:05:15 Arrakis emhttpd: spinning down /dev/sdo Apr 20 16:05:17 Arrakis emhttpd: read SMART /dev/sdo Apr 20 16:05:26 Arrakis nginx: 2026/04/20 16:05:26 [error] 308118#308118: *2216296 upstream timed out (110: Connection timed out) while reading upstream, client: 192.168.50.22, server: , request: "POST /update.htm HTTP/2.0", upstream: "http://unix:/var/run/emhttpd.socket:/update.htm", host: "192.168.50.158", referrer: "https://192.168.50.158/Main" arrakis-diagnostics-20260420-1615.zip

Thanks I didn't notice that I left stress on autostart ;)

Yeah, there was some issue with it

Yeah something was odd.arrakis-diagnostics-20260323-0809.zip

But we can transfer the license and tie it with TPM correct?

Just noticed that I have out of memory error, are you able to let me know what caused it? arrakis-diagnostics-20260323-0809.zip Was it because of "stress" application or there was something else?

I was looking where connection stops, I had to disable authentication for the tests. But now I stuck on this: It wolduld be nice to know if issue is with wopi or with collabora

Yup, there was something going on between Pangolin OIDC and opencloud, had to do some rules and voila:

I think my issue is on the pangolin side at the moment.

I have deployed a lot containers in my life, had some issues previously with some of them but this thing, I finally managed to do it and bum: You know what is funny? That I had fully working instance of opencloud, but I was going to add collabra, and I tought that this will speed it up lol. I give up, not sure what is wrong, I am doing everything step by step and every time there is something broken.

Did fresh install as it was not accepting ANY password, now I am facing this: I literally have no clue what is going with this thing, why is it so hard to set it up?

Well I put admin because mine something like "&R7$Tyklo56" was not working. I will try once again.

it does not accept the login/password combination, no matter what I put as a password :)

Weird, it is blank but I can see it in code.

Approach number 2: Did step by step, do I have to run init in this version? If I do this should be in documentation.

Thanks for your work, I did tried this few hours ago and after 30 min I dropped it. I will try again later once my sanity is back. Do I need need SWAG even if I have pangolin as reverse proxy?

Thank you I will have a look in upcoming days, I have deployed opencloud already so I think I will wipe it and use your script to get everything working as it will be painful to do everything alone.

I will have a look on, I thought env can be loaded from the file or I could simply add some extra things in the template. I moved from nextcloud AIO as it was overengineered and slow, opencloud is faster but looks like once I finish everything it wont be any less easier. Now as I've seen your post I can say, there will be no easy way to get it running unfortunately. Do I really need wopi and nginx? I am already using pangolin as reverse proxy, do I need another one running internally? Cant I just have opencloud + collabra?

I've edited the post. Yes I did fixed it by changing permissions, it looks like I need env to change e-mail settings, that is weird :D