Skip to content
View in the app

A better way to browse. Learn more.

Unraid

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

[Unraid 6.12.13] Drive in Cache Pool failed, came back, now Docker won't start...

Featured Replies

Good-day to All,

 

This is my first post and I'm going to try to offer as much relevant information as possible concerning the issue i've run into.

 

Running Unraid 6.12.13 and had what seemed like a pretty stable setup.

 

Back on November 17th, when I had some time off, I relocated the server in a new room. When I went to restart the server in its new location, I got multiple errors during boot-up that were stating something was wrong with the nvme / PCI-express devices. I snapped some pics with my phone and have attached one of the images to this post.  Among the error messages, there was the text "AER: Multiple Corrected error message received from 0000:09:00.0" and "PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)".

 

After many lines scrolled by with those messages, the server managed to start. I logged in to the WebUI and saw that one of the NVMe drives in my cache pool [2 NVMe drives total] was not seen by Unraid. Of course, nothing would launch. No Docker and no VMs.

 

I reboot the server twice and the same errors were seen during startup but strangely enough, Unraid saw both NVMe drives again.

 

I decided to run the BTRFS scrub on the cache pool and there were no errors. Despite this, Docker will not launch and VMs will not start.

 

I left the server running and just walked away for a month to take care of work and personal matters.

 

Just earlier this week, I had some time again to try to troubleshoot the server, so I logged in to the WebUI again and had Unraid generate the zip file with all the logs [attached].

 

As far as I can tell Unraid still sees both NVMe drives in the cache pool. Furthermore, there do not appear to be any errors with the NVMe drives based on the SMART tests I found in the logs, so I am really stuck as to what to do next.

 

Are there any command-line utils I could use to check the health / status of each NVMe drive in the cache pool? Should I take out the NVMe drives and check them in another system?

 

Thanks in advance!

 

PXL_20241117_063744914.jpg

tower-smart-20241216-1314.zip tower-diagnostics-20241216-1310.zip

Solved by JorgeB

  • Community Expert

New diags after array start please.

  • Author

@JorgeB:

 

started the server and started the array this time. Ran the diagnostics again. File is attached.

tower-diagnostics-20241223-0409.zip

 

in the syslog I found this line:

 

Dec 23 04:09:14 Tower kernel: BTRFS info (device nvme0n1p1): bdev /dev/nvme1n1p1 errs: wr 62078695, rd 573, flush 0, corrupt 14937, gen 614

 

which looks like there are both write and read errors with a number of corrupt blocks!!

 

Any advice on how to proceed / what to do next is most appreciated!

Edited by Dosage278
updated information

  • Community Expert

Run a correcting scrub on the pool and post the results.

  • 2 weeks later...
  • Author

@JorgeB Attached you should see a screen snippet after I ran the BTRFS scrub on the cache pool. It reports no errors found! This is very frustrating because at some level, Unraid recognized there are errors but now the test does not find any errors 😞

 

Any advice is greatly appreciated!

 

 

btrfs.scrub.results.part-01.2025-01-01_13-42.png

  • Author

Hi again and Happy New Year @JorgeB...

 

Appreciate all the guidance thus far. Today I followed the steps as outlined in the post you linked to and managed to clear the BTRFS stats. See attached screenshots. Listed the stats, cleared the stats, the same values were reported, reboot the server, listed the stats again, and finally the stats were cleared. Unfortunately, Docker engine still failed to start. I have attached diagnostics again [with array started!]... Going to look at the logs now to see if I can find anything obvious.

 

Question: how sensitive are connectors for U.2 NVMe drives? My cache pool is composed of 2 identical U.2 NVMe drives and this all began when I moved the server to another location. Put differently, should I go ahead and buy another cable and see if that solves the matter? If there is a problem with the U.2 cable / connector, why would Unraid report that it sees both cache drives?

 

 

btrfs.dev.stats.for.cache.pool.after.reset.png

btrfs.dev.stats.for.cache.pool.force.reset.png

tower-diagnostics-20250103-1659.zip

  • Author

@JorgeBThanks, that did the trick!

 

I guess whatever happened with the cache pool corrupted the docker image file...?? Based on what you said in the other thread [which is from 2018!], I take it that Limetech has still not adopted checking the BTRFS dev stats into Unraid natively? So the best practice is to set up that script which is outlined there?

 

Thanks again for all the help!

  • Community Expert
16 hours ago, Dosage278 said:

So the best practice is to set up that script which is outlined there?

Yep

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.