NVMe Drives Causing Stability Issues

[crspangenberg]

I have my server setup following SpaceInvaderOnes guides, I created a cache pool for my Appdata on an NVMe drive and for my VMs on an NVMe drive. This worked for awhile but I started notcing that my VM would stop running (go into the paused state) randomly, and then I would have to force stop the VM in order to start it back up.

 

Occasionally I would also see that my docker images would stop working and it would lock up Unraid entirely and I would be forced to restart manually. 

 

I have tried unplugging and reseatting each of the NVMe drives with no luck, I recently moved all of the data off of those drives, and onto a standard cache SSD that I was using for my downloads share, and everything seems to be functioning normally again.

 

Before I remove the drives, is there anything I could do to remedy the issue, or troubleshoot the issue?

 

My build is as follows:

Unraid 6.12.10

 

Any help would be greatly appreciated.

 

 

[JorgeB]

38 minutes ago, crspangenberg said:

I started notcing that my VM would stop running (go into the paused state)

AFAIK only to things can cause this, sleep being enabled in the VM OS, or running out of space in the storage where the vdisk resides.

 

You should also post the diagnostics after you notice the problem.

[crspangenberg]

Jorge, I appreciate the reply. 

 

Interesting, I would think that with either of these problems, they would continue to happen on the SSD as well, but they only occured on the NVMe drive. I will try to move the data back to the nVME drives and see if the problem persists. If they do I will try to download my diagnostics file. Although sometimes it locks up unraid completely. Is there a way to download the diagnostics when Unraid cannot be accessed via the GUI?

Edited July 2 by crspangenberg

[itimpi]

You can also get them from the command line as mentioned in the link.

 

The syslog in the diagnostics is the RAM version that starts afresh every time the system is booted.  You should enable the syslog server (probably with the option to Mirror to Flash set) to get a syslog that survives a reboot so we can see what leads up to the problem.  The mirror to flash option is the easiest to set up (and if used the file is then automatically included in any diagnostics), but if you are worried about excessive wear on the flash drive you can put your server's address into the remote server field. 

[crspangenberg]

Understood, I will enable the syslog server and get the diagnostics uploaded on the next crash.

[crspangenberg]

I performed a bios update, and moved the data back onto the drives, no issues for almost a week. Thank you for all of your help.