Disks read errors!! -SAS card at fault? Mobo? GPU?

[spdelope]

I have had a couple times this past week or so with major errors with my array where all disks attached to the SAS expander card would be unavailable. Woke up to an array that wasn't available and a full log file. 

 

This sort of happened when I replaced a 10GBaseT card with an SFP+ one. Not sure if related. A day or so later, it said one of my drives failed so I replaced and rebuilt the array. Then I had an unresponsive server. After some reboots and updating BIOS and ensuring BIOS settings are correct, I finally was able to get it booted and working when I pulled the GPU. So far so good.

 

So not sure if related to the GPU or not, or maybe its some sort of PCI lane issue with the mobo or power management with PCI. Or it could be the SAS card failing. Some help is appreciated.

 

I've attached logs from when it happened two times. At the end of the two logs are the read errors which continue almost infinitely.

logs relevent #2.txt logs relevent.txt

unraid-server-diagnostics-20240411-1209.zip

Edited April 11 by spdelope

[JorgeB]

Both times it was a HBA problem:

 

Apr 10 02:10:01 UnRaid-Server kernel: mpt2sas_cm0: SAS host is non-operational !!!!

 

Make sure the HBA is well seated, you can also try a different PCIe slot, also see below, to see if it helps with the PCIe errors being logged:

https://forums.unraid.net/topic/118286-nvme-drives-throwing-errors-filling-logs-instantly-how-to-resolve/?do=findComment&comment=1165009

 

[spdelope]

9 hours ago, JorgeB said:

Both times it was a HBA problem:

 

Apr 10 02:10:01 UnRaid-Server kernel: mpt2sas_cm0: SAS host is non-operational !!!!

 

Make sure the HBA is well seated, you can also try a different PCIe slot, also see below, to see if it helps with the PCIe errors being logged:

https://forums.unraid.net/topic/118286-nvme-drives-throwing-errors-filling-logs-instantly-how-to-resolve/?do=findComment&comment=1165009

 

 

Interesting that turning ASPM on seemed to make everything happy and work together. I also reseated the card for good measure. Thanks for the reply!