Frequent PSOD - PF Exception 14 in world NVMeComplWor

[qabrain]

Had ESXI 7 running on Dell OptiPlex 7090 for 6 months without issue. For the past 3 months has been PSODing at random times every 6 hours to 6 days. More frequently under load. I have attached a screenshot of the psod.  Any ideas?

[pmichelli]

We had a similar issue with one of our MX750c blades.  I am no expert in reading these, but from my calls with VMware, PCPU is physical. In our case it was the processor we had to get replaced.

We had 2 in the Blade and PCPU2 would be the one in the dump all the time. Dell asked us to swap the procs and sure enough CPU1 started failing, so they replaced the processor.

What does the iDRAC show? I would run the diagnostics from there as well

Is all your firmware up to date and compliant under the HCL ? 

[qabrain]

Thanks for the response. The Dell Diagnostics shows no issues.  I cannot flash the firmware on the drive since there is no support for doing it outside of windows.

[pmichelli]

Have you double checked that all the gear you have is compatible with ESXi 7.0.3? I have had issues in the past with vSAN when firmware and/or hardware was not officially supported.

What type of controller does that NVMe attach to? Apologies I have never used one. Trying to see if they use a BOSS or PERC card because those will have firmware for sure that you can patch.

If you have warranty, call Dell. They can help you figure out if the firmware is compliant or not. I've had good success working with their support teams

[Kinnison]

Hi,

I beg your pardon, but I'm afraid there isn't much to look for in terms of "compatibility", that DELL Optiplex 7090 system is commercially classified as a "Business PC" and its manufacturer does not appear (to me) to have declared it compatible with some version of ESXi. In the end, unfortunately it is always the same story, using an operating system that is not explicitly supported is always an unknown factor.

Then it would have to be said that the OP does not specify which exact system model it purchased, and even if its problems have emerged after any changes / additions to the initial configuration.

Regards,
Ferdinando

[qabrain]

Im not sure what controller type the NVMe attaches to.  In terms of compliance, i can say that the ssd and memory is not listed as being supported by ESXI 7, however, a number of people on my team are using this same exact setup without issue.

[Kinnison]

Hi,

So, If your contribution towards those who try to help you boils down to writing that: "however, a number of people on my team are using this same exact setup without issue" I sincerely hope you can solve your specific problem. As far as I'm concerned, the discussion ends here.

Regards,
Ferdinando

[qabrain]

Update - I am running 30 vms on this Server. I updated all vms, changing the disks from the SCSI controllers to SATA or IDE, and since making this change approximately 7 days ago, have not had a psod.

[pmichelli]

You've also just killed any and all performance from the disks by selecting those archaic controllers.