I had two disks go disabled due to read errors overnight.  (Yes I have notifications turned on and this is the first notification I received of any issue with them.)  For reference, I had a previous *issue* with most of my disks attached to my HBA SAS card a while back but only one disk went disabled due to write failure.  The two disks that have become disabled are the only two left on the HBA SAS card at the moment (I thought ahead and tried to keep the disks that were getting written to constantly directly connected to the motherboard's SATA ports).

 

My guess is that this is directly related to the HBA SAS card (I have ordered another one and it should be here soon, I will also be working toward a custom cooling solution for the new card).  

 

I shut down the server. Moved the breakout cable to a different port on the card and the disks have reappeared as connected for the time being. Here is a copy of the diagnostics with the short SMART tests ran on all the SATA connected devices.  I will be running the longer tests (with the array disabled) while I am at work and will report back when I have those results.)  From what I can tell from the *Attributes* the disks look good; it is either a cable issue or that HBA SAS card.

 

I have included diagnostics before the shutdown and cable move. I have also included another set from after the disks came back online and I ran the short SMART tests on all the SATA connected drives (including the breakout cable ones).

 

Edited May 1, 20251 yr by HeliusSol

Apr 16 02:07:28 SpiderWeb kernel: mpt3sas_cm0: SAS host is non-operational !!!!
Apr 16 02:07:29 SpiderWeb kernel: mpt3sas_cm1: SAS host is non-operational !!!!
Apr 16 02:07:29 SpiderWeb kernel: mpt3sas_cm0: SAS host is non-operational !!!!

 

HBA problem, make sure it's well seated and sufficiently cooled, you can also try a different PCIe slot.

@JorgeB Just wanted to follow up. I assume this is a heat issue or something else at this point. Just checking, that you didn't see anything related to Disk Failures beyond that SAS issue?  Not sure what I'd be looking for myself other than stuff in the attributes tab (or the files for those disabled disks) that looks "off"...

3 hours ago, HeliusSol said:

that you didn't see anything related to Disk Failures beyond that SAS issue? 

I didn't.