500 error - Cannot access UI externally or from Safe Mode

[Jambo]

homeserver-diagnostics-20230908-1132.zip

 

Diagnostics attached. I set the "Autostart Array on Startup" to false so I could add a new drive and immediately the UI crashed and only responded with error 500.

 

I restarted the system and on startup, the array did not start, and the UI is still responding error 500.

 

I've tried the following commands to no avail:

 

• /etc/rc.d/rc.nginx reload
• /etc/rc.d/rc.php-fpm restart

 

Any help?

 

Logging in via the SAFE Mode with GUI, it also fails, with just a white screen showing in Firefox.

Edited September 8, 2023 by Jambo
Clarify that array now starts but UI inaccessible.

[Jambo]

Looking through the syslog I can see:

 

Sep  8 11:29:16 HomeServer crond[1925]: exit status 255 from user root /usr/local/emhttp/plugins/dynamix/scripts/monitor &> /dev/null
### [PREVIOUS LINE REPEATED 1 TIMES] ###
Sep  8 11:30:32 HomeServer emhttpd: error: cmdStart: missing csrf_token
Sep  8 11:31:16 HomeServer crond[1925]: exit status 255 from user root /usr/local/emhttp/plugins/dynamix/scripts/monitor &> /dev/null

 

Not sure if they're relevant?

[JorgeB]

Post new diags after booting in safe mode.

[Jambo]

Thanks for the interest, new diags below. GUI Mode shows blank white screen on Unraid's own firefox.

homeserver-diagnostics-20230908-2320.zip

[JorgeB]

I don't see a reason for the GUI not working, try booting the server with a different stock flash drive, no key needed, just to see if it works and see if it's a config issue.

[Jambo]

Yep, that works. I assume it's no help but diags below just in case.

 

How do I get to the unraid connect flash backup without the server GUI? Or do I have to find the issue in the config?

tower-diagnostics-20230909-0229.zip

[JorgeB]

I would backup current flash drive, recreate it and then restore just the bare minimum, like super.dat and pools folder for the assignments, the key file, and the docker user templates folder, then boot to confirm it's still working, and if yes either reconfigure the rest of the server or restore a few files at a time from the backup (config folder only) until you find the culprit.

[Jambo]

I'm genuinely shocked that the answer is to start from scratch and rebuild to solve this issue.

 

Also, the Unraid connect backup is not accessible if you don't have a working server? What's the point of it?

 

So the key files to start with are:

 

/config/super.dat

/config/Plus.key

/config/pools/

/config/plugins/dockerMan

 

I'll also change the disk.cfg to prevent auto-start and just check the UI before continuing.

 

I'm guessing after that, copy shares and all of plugins over? Would you go as far as copying over all the config after that?

[itimpi]

1 minute ago, Jambo said:

Also, the Unraid connect backup is not accessible if you don't have a working server? What's the point of it?

You can get to it from the forum - no need to have a working server.

[Jambo]

So I've got it running and working again, Selectively and slowly putting the plugins back has narrowed it down to either:

• Dynamix SSD Trim
• intel-gpu-top

No idea which but whatever. Can't test now as it's doing a 16tb disk clear

 

8 hours ago, itimpi said:

You can get to it from the forum - no need to have a working server.

Thanks mate - didn't see the option the first time I ventured in there!

[SimonF]

21 minutes ago, Jambo said:

So I've got it running and working again, Selectively and slowly putting the plugins back has narrowed it down to either:

• Dynamix SSD Trim
• intel-gpu-top

No idea which but whatever. Can't test now as it's doing a 16tb disk clear

 

Thanks mate - didn't see the option the first time I ventured in there!

Trim is in os from 6.11 so wou!d be likely issue.