Hi all

My son's GPU in his computer died, so seeing that I had a GPU in my unRAID server that I haven't been using recently, I decided he could use it. I unbound it in vfio, shut the server down and removed the GPU.

When I restarted the server, my VM's wouldn't start automatically, even though the VM which had been using this GPU, wasn't set up to start automatically. Anyways, I deleted that VM and tried again. The server will boot, and at first I can access the GUI in a web browser, but not very long after access stops. When this happens I have no choice but to do an unclean shutdown. This happens each and every time.

I've been able to boot into GUI Safe Mode, and have been able to download the Diagnostic file, which I'll add to this post. When I try to start the array, it hangs there.

I was running 7.1.4 but early on while trying to fix this, I rolled back to 7.1.2.

Here is the Diagnostic file: tower-diagnostics-20250713-1239.zip

Any help would be very much appreciated!

EDIT: I have to go out for a few hours, so before leaving I booted the server the normal way and left. I noticed in the Tailscale app on my phone that the server was connected, so at least the server got that far and started the Tailscale service, but my excitement was short lived coz a few moments later it was no longer connected. While it was connected I did try to access the server, but it didn't work.

Edited July 13, 2025Jul 13 by HomerJ
Added additional info

UPDATE: I just got home and went straight to trying to fix this. I removed the USB drive and using another computer I recreated the boot drive and then transferred back the original CONFIG folder.

First boot, the server IS online, but I can't access it. According to unRAID Connect, it's up and running.

Even according the unRAID Connect, there are issues, because there is a lot of info it is unable to show.

I have rebooted my router, which made no difference. I am also unable to access any of the shares.

Thanks to the search function and the great help that is available here, I've been able to work out how to get a syslog file! I haven't looked at it yet, unfortunately it's not very likely that I'll find anything that makes much sense to me lol but I will have a look anyways once I finish this post. The file is attached:

syslog

I still haven't had any success. I noticed issues reported with one of the drives. It's getting late here, so I'm going to leave this now until tomorrow. Tomorrow's plan is to replace the SATA cable to the drive that the error was attributed to, and then we how it goes.

If I understood correctly it works in safe mode? If yes, it suggests a plugin is the problem.

21 minutes ago, JorgeB said:

If I understood correctly it works in safe mode? If yes, it suggests a plugin is the problem.

Yes and no. It will boot into safe mode where I can access the GUI, but when I go to start the array, it hangs.

Did you see any issues in the syslog? I thought I'd seen that there was an issue with one of the drives, which is why I'm planning on replacing the SATA cable to it tomorrow morning.

1 hour ago, HomerJ said:

Did you see any issues in the syslog?

No, but the syslog is before array start.

Boot in safe mode, start the array, and then get the syslog from the CLI:

cp /var/log/syslog /boot/syslog.txt

Thank you for your reply!

So tomorrow I will boot into safe mode, then start the array. It will then cease to be responsive so I then do an unclean shutdown. Then I'll boot into safe mode again and I then type the command you suggest in unRAID 's command prompt, right?

Sorry that it night seem that I'm seeking help with wondering that seems easy, I just want to be sure that what your suggesting is what I think it is.

39 minutes ago, HomerJ said:

It will then cease to be responsive so I then do an unclean shutdown.

The syslog must be saved before rebooting from the CLI.

The syslog in the diagnostics is the RAM copy and only shows what happened since the reboot.   It could be worth enabling the syslog server to get a log that survives a reboot so we can see what happened prior to the reboot. The mirror to flash option is the easiest to set up, but if you are worried about excessive wear on the flash drive you can put your server’s address into the Remote Server field.

10 minutes ago, itimpi said:

The syslog in the diagnostics is the RAM copy and only shows what happened since the reboot.   It could be worth enabling the syslog server to get a log that survives a reboot so we can see what happened prior to the reboot. The mirror to flash option is the easiest to set up, but if you are worried about excessive wear on the flash drive you can put your server’s address into the Remote Server field.

This is what I've done. I setup the syslog server, while in safe mode. I then tried to start the array but it hung. I forced the server to shut down. I then took the USB stick to another computer and copied the syslog file, which I shared a couple posts up. Wouldn't this mean that the syslog file I've shared will have the information needed to identify the issue?

I just replaced the data cable to the hard drive which I thought had an issue from what I read in the syslog, but it's still not working.

Right now I'm booting again into Safe Mode with the hope that I can remove some Plugins to see if any of them are the issue.

The server has had a couple restarts so I'm hoping the latest syslog will have some info, which I've attached here.

syslog (1)syslog-previous

Ok, I've done as suggested. I was lucky coz I booted into Safe Mode, started the array and created the syslog file, then very soon after I could no longer access the GUI.

I had to change the name, but it is the right file.

By the way, I even tried recreating the USB drive and only transferring files related to my shares in the CONFIG folder and the server still wouldn't run.

syslogfile.txt

Edited July 14, 2025Jul 14 by HomerJ
Additional info added

2 hours ago, HomerJ said:

I was lucky coz I booted into Safe Mode, started the array and created the syslog file, then very soon after I could no longer access the GUI.

SO the syslog was copied before the GUI stopped responding? I understood it stopped during array start. Enable the syslog server and post that after it happens again.

I started the array, then I typed

cp /var/log/syslog /boot/syslog.txt

5 to 10 seconds later the server was no longer contactable. I pulled the USB stick out and got the file from it.

The problem happened after the syslog was copied, so there's nothing relevant there.

Ok, thank you JorgeB and itimpi for trying to help me. Yes I suspected that the issue occurred after it created to file. I'm some understand how to have it create the file AFTER the issue occurs; I thought this was what 9 had done when I setup the syslog server which provided me with the original syslog files.

Anyways, I've been struggling most of today trying to fix this in between doing other stuff at home, and decided to simply record the disk assignments and then recreate the unRAID server. So far it's been running as it should for the last 3 hours. It's just a pity that I wasn't able to figure out what the issue was. I had thought that the cause may have been the fact that I had removed a AMD GPU while the Plug-ins for it were still present and running, but even after I removed their folders in the Config folder on the boot USB, the issue was still there.

The good news though is that I'm slowly moving forward to having the server up and running again as it should.