ESXI PSOD

[beire]

Hey guys,

 

I've been running an ESXI host with unraid as VM for the last few months.

It has been rock solid, until today when i noticed nothing was working anymore.

 

Turns out the host had a PSOD.

 

I have attached two screenshots of the screen. It occurred two times. The second time only after 10-20min of uptime, whilest copying some files.

 

I am now running bare metal unraid doing parity check. It seems fine until now, i copied the same file again with good result.

 

This is my hardware setup:

 

intel xeon e3-1240 v2

supermico x9scm-f

16gb ram

supermicro SAS2LP-mv8 with hack

 

Is there enough info in the screenshots to help debugging this issue. I see the e1000 adapter is coming up on both screendumps. But actually i have no clue where to start.

 

Any help is appreciated!!!

 

 

[BobPhoenix]

Check that your SASLP-MV8 that you have passed through still has the pciPassthru0.msiEnabled = "FALSE" in the configuration for your VM.  If you made hardware changes to the server that setting can easily get changed.  Especially if you had to change the VM settings to pass it through again the msiEnabled line will get deleted.  Beyond that you might have to seek help on the VMWare forums if nobody else has a better idea.

 

I've had a SASLP-MV8 on pass through to a VM for 6 months without a problem myself and it is only when I change hardware and reconfigure the VM that I get PSODs.

[beire]

Thanks for your reply BobPhoenix.

 

The hack was still present in the configuration. I double checked.

 

Seems that, after some googling, i found this to be a known issue with ESXi 5.1 and 5.5 regarding the E1000 adapters.

 

It should be resolved using the vmxnet3 adapter instead. Upgrading to 5.5 Update 1 should solve the issue too.

 

I'll give it a few days with the vmxnet3 adapter and see. I'll update to UPDATE1 when i have the time.

[BobPhoenix]

Thanks for your reply BobPhoenix.

 

The hack was still present in the configuration. I double checked.

 

Seems that, after some googling, i found this to be a known issue with ESXi 5.1 and 5.5 regarding the E1000 adapters.

 

It should be resolved using the vmxnet3 adapter instead. Upgrading to 5.5 Update 1 should solve the issue too.

 

I'll give it a few days with the vmxnet3 adapter and see. I'll update to UPDATE1 when i have the time.

AH.  I enabled the vmxnet3 adapter from the start that way I get 10Gbs transfers between VMs.  Also I'm still on ESXi 5.0 which might also explain the difference.