[Help] UNRAID acting erratically

[Astatine]

So, I was moving houses and couldn't pay attention to my UNRAID server for quite some time. I have since replaced a corrupt USB drive. Transferred my key to a new USB. 
It was working properly for some time but soon after that I started noticing that the server would "freeze" as in the docker services would be running but a couldn't access server using the IP. 

I also noticed that during boot there are these number list (my guess memory block list) which reads so: 

I don't understand what this is. Then the situation worsened. Now, docker containers would just randomly stop working and when tried to restart them, I would receive popup either saying "403" or "Execution error". I then either had to restart the docker service or the whole server to be able to get them to work. My guess was that the docker file is corrupted. But I couldn't still act to fix that due to life. 

 

Then the whole server started freezing. I couldn't access it over the IP, same for docker containers until I restart the machine which resulted in a ton of unclean shutdowns. 

 

And as of yesterday, after restart, the docker containers are only working for 5-10 minutes and my unraid machine name went back to "Tower" which according to this could be config issue: Server name changed from "N2" to "Tower" - General Support - Unraid

Also, I am no longer able to access the server over https://[IP] but http://[IP] is showing the login screen but after entering the credentials, it doesn't take me in. Just keeps redirecting to login screen. 

 

I convinced that something is seriously off in the /config in the USB. 

 

Can someone help me track down the issue and fix it? or is it better to start from scratch? 

If so, when I do 'New Config' and assign the devices to their current purpose, will my data be retained? both array and cache? I understand that new config will definitely rebuild the parity.

 

I try to somehow get the diag. report from the server directly and post it here soon

Here are the diagnostics: tower-diagnostics-20240119-1839.zip

Also, when I clicked 'Download' on the diagnostics page, I saw the following error on the server cli screen: 

 

Please help. It is getting very frustrating to deal with!!!!

Edited January 19 by Astatine
Added diagnostics and error screenshot

[JorgeB]

Jan 19 15:38:09 Tower kernel: macvlan_broadcast+0x10a/0x150 [macvlan]
Jan 19 15:38:09 Tower kernel: ? _raw_spin_unlock+0x14/0x29
Jan 19 15:38:09 Tower kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]

Macvlan call traces will usually end up crashing the server, switching to ipvlan should fix it (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)), then reboot.

 

10 hours ago, Astatine said:

Also, when I clicked 'Download' on the diagnostics page, I saw the following error on the server cli screen: 

Docker image is corrupt, delete and recreate.

[Astatine]

Quote

Macvlan call traces will usually end up crashing the server, switching to ipvlan should fix it (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)), then reboot.

I previously had issues with docker containers, thread linked. Which was solved by switching to macvlan.

 

If I go back to that, can you suggest how to avoid the DNS issue again?

I'll try switching back to ipvlan this evening and see if it makes things better

Edited January 20 by Astatine
added more info

[itimpi]

9 minutes ago, Astatine said:

Which was solved by switching to macvlan.

If you want to continue using macvlan then you need to disable bridging on eth0 as mentioned in the Release Notes.

[Astatine]

Quote

If you want to continue using macvlan then you need to disable bridging on eth0 as mentioned in the Release Notes.

Understood. I'll go through the release notes to understand the new way to do bridging. 

 

Anyways, what should I do with the potentially corrupt config? My server renamed itself to 'Tower'. ident.cfg file lists name as 'Tower'. I don't know what else might have been reset back to default. Do I need to start with 'New Config'? 

[JorgeB]

I would just reconfigure what changed, just change the name back for now.