UNRAID crashed

[tr3bjockey]

Everything was going fine except that all of a sudden, my drives all disappeared from the array.  4 of the rust spinners are connected to the add-on card (see signature) 2 of the SSD's connected to MB.  Watched a movie from plex yesterday and didn't have any issues.  Copied more stuff to the drive and right after the copy job...I get the screen below. 1st screen...everything is gone including the unraid thumbdrive.  Then a few seconds later, I see the drives come back, but they are all unassigned.  Then they all disappear again, and then again show up as unassigned.  So I go into the diagnostics to make a diagnostic log file for you guys and I get an error trying to do that.

 

This week I replacement my 16GB of ram which weren't working with the Patriot Viper memory sent to me by patriot as part of a lifetime warranty exchange.  Yup, I bought the ram in 2008 and they sent me brand new sticks to replace them under warranty.

 

Edited April 19, 2022 by tr3bjockey
change the order of pics

[tr3bjockey]

Downloaded memtest86 and running it off of a usb slot...  So far no memory errors on pass 1/4.  Will post a pic of the finished test.

 

Just don't understand why I can't catch a break.  LOL

 

Sorry admins for being such a pain in the butt.  I know I've had 3 different issues in the past two weeks.  On the positive side, I think it ran fine for over a year without issues.  (using it about 16 hours a week)

[Squid]

Did the diagnostic collection ever complete?

 

Did a file get created on the flash drive (logs folder) of the diagnostics?  Upload it if it exists. (It's date & time stamped)

 

Otherwise, your recourse pretty much is a reboot.  For that directory "state" or var.ini to not exist really implies that something went through and really trashed the OS folders.  I have never seen anything ever do that.

[tr3bjockey]

Perfect timing...didn't think it created one because of the error...but it did on the flash drive.  Thanks Squid.

tower-diagnostics-20220419-1407.zip

[Squid]

This is my first guess

/tmp
total 0
drwxr-xr-x 3 nobody users  60 Apr 19 12:54 Transcode

You have (plex I imagine) transcoding to RAM.  No problems with that.  My thought is that in the Plex template you've got /transcode mapped to /tmp and within Plex you're telling it to transcode to /transcode/transcode.  Shouldn't be an issue per se (that I can think of), but I can see tons of very weird things happening if /tmp ever gets deleted by an App.

 

Your only solution right now is a reboot.

 

5 minutes ago, tr3bjockey said:

didn't think it created one because of the error.

Yeah, it's a not very well known feature that even if the GUI crashes during diagnostics collection that the file still gets generated and stored on the flash.

[tr3bjockey]

I was not watching plex at the time this happened.  I know it still runs in the background. FYI

 

Did you see any errors related to every single drive including unraid usb disappearing?  Would the system do this if the USB was going bad and disappeared from the system or would it behave completely differently if that were the case?

 

Still on pass 1/4 for memtest.  30 minutes in and no errors...  I also noticed I was getting a few errors on the parity drive related to bad cable.  I switched out all the sata cables with new cables from https://www.amazon.com/gp/product/B01IBA335Y/  last week.  With one of the admin's help, I had an SSD disappear and reappear and it was corrupt.  Got that added back in to the cache pool as a raid 1.  So far (besides the data loss I had from a partition getting corrupted on my rust spinner), it looked like it was stable.

 

Do you think I need to do the full 4/4 passes of memtest or by now we should have seen something, before I reboot with my unraid usb?  BTW, thanks for the lightning fast response!

Edited April 19, 2022 by tr3bjockey

[tr3bjockey]

23 hours ago, Squid said:

This is my first guess

/tmp
total 0
drwxr-xr-x 3 nobody users  60 Apr 19 12:54 Transcode

You have (plex I imagine) transcoding to RAM.  No problems with that.  My thought is that in the Plex template you've got /transcode mapped to /tmp and within Plex you're telling it to transcode to /transcode/transcode.  Shouldn't be an issue per se (that I can think of), but I can see tons of very weird things happening if /tmp ever gets deleted by an App.

According to an article on the unraid forum, it's configured correctly.  Do you think I should change it in some other way?

"There are two tweaks to be made in order to move your transcoding into RAM.  One is to the Docker Container you are running and the other is a setting from within the Plex web client itself.

 

Step 1:  Changing your Plex Container Properties

From within the webGui, click on "Docker" and click on the name of the PlexMediaServer container.  From here, add a new volume mapping:

 

/transcode to /tmp

 

Click "Apply" and the container will be started with the new mapping.

 

Step 2:  Changing the Plex Media Server to use the new transcode directory

Connect to the Plex web interface from a browser (e.g. http://tower:32400/web).  From there, click the wrench in the top right corner of the interface to get to settings.  Now click the "Server" tab at the top of this page.  On the left, you should see a setting called "Transcoder."  Clicking on that and then clicking the "Show Advanced" button will reveal the magical setting that let's you redirect the transcoding directory.  Type "/transcode" in there and click apply and you're all set.  You can tweak some of the other settings if desired to see if that improves your media streaming experience."

[Squid]

IDK.  You just had a weird error, and that was my first guess on what the cause was, and what you were seeing is technically impossible to happen