Hello,

 

Since August 2nd, I have been having an on-off issue since I have replaced my Google Router with an Asus Router (ASUS RT-AX58U). This does not happen to another other device but it has happened once with my 2nd UNRAID server. I hope the Asus router is not the culprit, I had hoped that UNRAID would be more resilient at the network level. My 2nd UNRAID server has been offline since the issues started happening. I also added an additional NIC in case it was the motherboard.

 

The UNRAID GUI would be unreachable but I would still be able to ping the server, my PLEX and other dockers are also unreachable via webpages. I have initially stopped all the dockers except for Plex, Cloudflare Tunnel, and Vaultwarden which is when it lasted over 24 hours. I started more dockers and was finally able to reproduce it after I started more dockers.

 

Since UNRAID is unreachable, I had to restart the server and the logs would be wiped on the reboot. I enabled 'Local syslog server' under 'Syslog Server' but there is so much logs. Would anyone be able to tell me which log to upload from there and possibly help me understand what is happening?

How are you assigning the IP address for your Unraid server--- DHCP or an assigned static IP address?  

 

If you have assigned a Static IP address on your Unraid server, have you reserved that address on your new router? 

33 minutes ago, Frank1940 said:

How are you assigning the IP address for your Unraid server--- DHCP or an assigned static IP address?  

 

If you have assigned a Static IP address on your Unraid server, have you reserved that address on your new router? 

The IP address is static on both ends - from UNRAID and the Asus Router. I am not sure if I can share the local IP here.

https://en.wikipedia.org/wiki/Private_network

 

Assuming you have assigned IPs that are allocated for private networks, and you haven't forwarded any ports to them from your router, then nothing outside your LAN can access them.

 

 

UNRAID:

 

Asus Router

 

Edit: I also compared MAC Addresses to make sure I was not misidentifying the UNRAID server with the second UNRAID server.

Edited August 8, 20241 yr by aysonohbata

Setup the syslog server using the 'Mirror to Flash' setting.  Then the next time it occurs, shut the server by pressing the power switch for approximately one second.  That should shut the server down cleanly.  Then pull the flash and get the file from the   /logs  folder and up load it in a new post. 

Edited August 8, 20241 yr by Frank1940

I did but its just that there is so much logs. The folder names there do not to match the date I enabled it.

logs.zip

I download your logs.zip file and it was not in there.  Your Syslog Setup should look like this:

 

And the  /logs   directory/folder should then contain this file:

 

That is the file that we are interest in seeing if you can't get the diagnostics file after the problem occurs.  If the GUI returns on its own after the problem, get the diagnostics file and upload it as well.  (The diagnostics file is useless if you have to reboot the server to get the GUI back!)

 

I see, I thought I had the correct Syslog Setup. Thank you for the screenshot.

 

Quote

(The diagnostics file is useless if you have to reboot the server to get the GUI back!)

 

That is the problem but with Syslog Setup from your screenshot it should save on the USB right?

3 hours ago, aysonohbata said:

That is the problem but with Syslog Setup from your screenshot it should save on the USB right?

Yes.

I think I have found out the problem before it happens and was able to prevent it. Lidarr has been grabbing a lot of files lately so Plex has been consuming a lot of RAM running Sonic Analysis. The RAM has been sitting at 99% and the UI started not responding. I was somehow able to stop Plex and the UI was then more accessible and I was able to navigate through it smoothly.

 

The RAM being filled would explain why the UNRAID is the only one inaccessible despite the router change. I guess a coincidence?

It also explains why it is pingable. If it was a problem at the router level, it shouldnt be pingable.

I am going to plug back in my second UNRAID to see what's going on there.

 

Is there a way to set up a limit on how much RAM a docker can use? Or should I let the problem happen again?

I guess not. I started more dockers and the UI became unresponsive then I got the usual message.

The server at 192.168.86.251 is taking too long to respond.

 

I managed to grab a screenshot before it kicked me out and hopefully the syslog is correct now. I had to manually turn it off to grab it.

 

syslog