UNRAID SERVER FREEZES | WEBUI FAILS TO LOAD | CONSOLE PARTIALLY UNRESPONSIVE #UNRAID_VERSION_6.12.8

[tjsyl]

X9DRi-F 2xE5-2690v2 160GB DDR3 15 drives with 0 errors.  

 

I was running 6.12.6 and lost my "MAIN" display, other tabs would load but "MAIN" wouldn't show the drives.  After 20+ days since the last reboot and seeing a couple posts about 6.12.8 fixing that issue I figured why not, I had already updated my X11DPH-T server and it has been smooth for a few weeks.  This issue seems to have started a few days after updating.  The Web UI and SMB shares become unresponsive but I can still pull up my console via IPMI or directly, I can type the user name but timeout before it ever shows the password prompt.  After trying that for a few times it becomes unresponsive and it may or may not drop a line when hitting return.  I was in the middle of a plex stream when it happened this time and the array was in process of a parity check after the last occurance (12TB x2 parity).  I can also see via the LED's on my LSI SAS controllers that the drives look to still be busy with what I assume to be the parity check.  I have yet to try pinging but If it happens again I will see if I get a reply.  

 

The system seems to take input when instructed to shutdown gracefully (FROM IPMI) but it complains about the hung process id 4597 (See Screenshot).  Is there anything in the diagnostics that can tell me what process that was/is?

 

I haven't made any changes to my bios settings but I am attaching some screenshots in an effort to verify if something is not playing nice with 6.12.8.

 

 

ur0-diagnostics-20240316-0118.zip ur0-syslog-20240316-0816.zip

[JorgeB]

Enable the syslog server and post that after a crash.

[tjsyl]

So far I've went down a rabbit hole and found the PLEX DB had some errors.   Full rebuild of the Plex DB via (https://www.reddit.com/r/PleX/comments/z7i4va/repair_a_corrupted_database_on_unraid_updated/) this helpful guide, I couldn't get ChuckPaPlex's script to play nice with UR, I think I may have been executing it from the wrong directory.  After manual full repair I am 80% through the parity check and no issues.

Just kidding. I was checking on it as I was typing this and I see something is running a muck on the RAM (98% of 160GB).  I had UNMANIC set to use the ramdisk long ago, (/tmp/xxxxxx) but I am 99% sure when I added the 4- 1TB ssd's (4-6 months ago?) changed it to one of the 2, 2TB cache pools I have set up.  I noticed Unmanic running out of space on very large Linux ISO's.   

 

I had the syslog writing to my other UR server and it looked useless, after working on some stuff for work I came back to it and noticed the server was responsive and it had the syslog had jumped from 14kb to 80+kb.  Apparently Unmanic was running a muck.  It would seem I don't have enough patients to wait it out the last few times this happened and took action before UR fixed itself.  For now I shut down Unmanic and disabled autostart, will try to see what's going on with that container.  Maybe move it to another server. 

 

Anything else helpful in this log?  I know exactly what the "smb_panic" was about.   I also omitted a few things *****.  Not sure why the tab open in chrome on my phone feels the need to log back in every 15 minutes but, meh..    I think all is well now but let me know if anything looks strange please.  I will disable writing the syslog to the flash drive for now but keep it mirrored to my other UR server.

 

syslog-10.xxx.xxx.xxx.log

[JorgeB]

10 hours ago, tjsyl said:

Anything else helpful in this log?

Nothing else I can see other the the mentioned UNMANIC issues.