Unable to connect or log into tower

[unTECH]

So, aside from occasional SMB access issues that would require a reboot, my tower has been running fine. Today, I had my shares become inaccessible, so I went to reboot it... Can't connect to the web GUI. Can't ping the tower. Connect a monitor and keyboard to the tower; was able to type in username "root", but when I hit enter, it goes to a new blank line; whatever I type shows up, but it isn't taking it as a password. 60 seconds later it times out and the CLI login resets, asking for username.

 

Since I can't connect to the web GUI and can't log into CLI... is there anything I can do? Or do I just need to do an unclean restart and hope the array isn't impacted?

 

Current version is unRAID 6.11.5

[itimpi]

A short press of the power switch on the server is meant to initiate the standard Unraid shutdown sequence (as though you used the Shutdown button in the GUI).

[unTECH]

Alright, so that did gracefully shut it down [I didn't realize it would stop the array, I thought that had to be done before sending for shutdown] -- unfortunately, after starting it up again, it seems to be in the same state. Can't ping it at the IP it lists, same behavior with not taking a password and timing out... Is my only option to try rebuilding the flash drive at this point, since I can't log in at all?

[JorgeB]

You can download and post the diagnostics to see if there's something visible.

[unTECH]

7 minutes ago, JorgeB said:

You can download and post the diagnostics to see if there's something visible.

Are the diagnostics files stored anywhere on the flash drive that I can access from another machine? I'm not even able to log into the CLI at this point [I didn't see anything listed in the link about what to do if CLI is inaccessible]

[JorgeB]

You can remove the flash drive and connect to a PC, they will be in the /logs folder, it's all in the link if you click it.

[unTECH]

On 6/14/2023 at 9:37 AM, JorgeB said:

You can remove the flash drive and connect to a PC, they will be in the /logs folder, it's all in the link if you click it.

Ahh, sorry, I thought /logs was the full logs without anonymization. Here you go!

 

---

Edit: removed diags because I realized they weren't sanitized.

 

Edited July 7, 2023 by unTECH

[JorgeB]

No link detected on bond0, which is your main interface, there's is link on bond1, swap cables or make the other NIC eth0, you can also just delete/rename /config/netwrok.cfg and network-rules.cfg and reboot, it will default to a bond with all NICs and DHCP.

[unTECH]

Hmm... So, turns out the main cable had slightly wiggled out of the port on the switch somehow. Swapped that out, now I see activity on all 3 interfaces from the back of the machine. However, I still can't ping it, connect to the web GUI, or log into the CLI... Going to try renaming the network files you mentioned

[JorgeB]

16 minutes ago, unTECH said:

Going to try renaming the network files you mentioned

If this doesn't work post new diags.

[unTECH]

Ok, so, interestingly... I tried just renaming network-rules.cfg, and it would recreate that file but still not work. Renaming both, it creates a new network-rules.cfg, but not network.cfg. It won't get a proper IP address [seems to default to 169.254.221.41, which is not in any of our scopes], but I can log into the CLI after renaming both. Which is good that it did, because it didn't automatically do diags with later reboots, so I was able to run the command.

 

 

---

Edit: removed diags because I realized they weren't sanitized.

Edited July 7, 2023 by unTECH

[JorgeB]

6 minutes ago, unTECH said:

but not network.cfg

That's normal, it would be created when you make a change.

 

Jun 14 06:54:33 Discovery  dhcpcd[1406]: br0: soliciting a DHCP lease
Jun 14 06:54:38 Discovery  dhcpcd[1406]: br0: probing for an IPv4LL address
Jun 14 06:54:44 Discovery  dhcpcd[1406]: br0: using IPv4LL address 169.254.221.41

This is usually a router problem, it's not leasing an IP address.

 

 

[unTECH]

2 minutes ago, JorgeB said:

That's normal, it would be created when you make a change.

 

Jun 14 06:54:33 Discovery  dhcpcd[1406]: br0: soliciting a DHCP lease
Jun 14 06:54:38 Discovery  dhcpcd[1406]: br0: probing for an IPv4LL address
Jun 14 06:54:44 Discovery  dhcpcd[1406]: br0: using IPv4LL address 169.254.221.41

This is usually a router problem, it's not leasing an IP address.

 

 

Hmm... There's a DHCP reservation for the server, must be hitting something there. I'll investigate the network side and follow up.

[unTECH]

So, update: Had some network issues take out a Windows server, so several changes were made; I was able to get the server accessible via the webGUI again.

Now, this is great, but there's still the CLI login issue. Put in the username, it goes to a new line, but isn't actually taking the password; whatever I type shows in plain text. After 60 seconds it times out.

I did find that if I boot the server in GUI mode, I can log in and do what I need to do, so it's not a huge problem... But still seems like a bug. I've now updated to 6.12.1 and it still does it.

[JorgeB]

3 minutes ago, unTECH said:

Now, this is great, but there's still the CLI login issue. Put in the username, it goes to a new line, but isn't actually taking the password; whatever I type shows in plain text. After 60 seconds it times out.

Not sure what it could be, but it works for me, it's not a general bug.

 

[unTECH]

Just now, JorgeB said:

Not sure what it could be, but it works for me, it's not a general bug.

It's definitely weird. I did notice that when I removed the network config files it would work, but then once I set the network up again, it would stop working... I guess I'll just ignore it since it isn't really causing an issue. Thanks for the help