System Crashes due to SMART tests

bswann93

Cadet
Joined
Nov 3, 2019
Messages
7
Hello, just to preface I am still somewhat of a noob with TrueNAS.

I recently updated from Core ver. 12U4 to 13U4. Ever since then anytime my extended offline SMART task runs my entire system crashes. sometimes some of the SMART tests on my drives finish and some of them fail. when I run a smartctl -a on a drive that "failed" the test the error it gives for it is "interrupted (host reset)". the most recent time this happened 5 of my drives were removed by the system due to errors resulting in data corruption. I booted the system and the drives came back without issue, and as far as I can tell my data is fine. after checking the smart tests for all 5 of the drives 3/5 passed with no errors and the others gave that same interrupted (host reset) error.

Does anyone have any idea what could be causing this?
 

Redcoat

MVP
Joined
Feb 18, 2014
Messages
2,925
Does anyone have any idea what could be causing this?
It would help if you provide the information detailed in the Forum Rules (red link at the top of the page) otherwise we can do nothing but guess.
 

bswann93

Cadet
Joined
Nov 3, 2019
Messages
7
It would help if you provide the information detailed in the Forum Rules (red link at the top of the page) otherwise we can do nothing but guess.
Apologies I had neglected to read it before posting.

TrueNAS Core 13.0-U4

supermicro X9DRH-7TF
2x Intel Xeon E5-2697 v2
256GB DDR3 ECC Memory
12x 8TB Seagate Iron Wolf
configure in two raidz2 vdevs
drives are connected to two different ASMedia PCI SATA controllers
PCI Intel I225-V 2.5GB NIC
TrueNAS is a VM hosted in WMware 7.0 - VMWare tools is installed on TrueNAS VM
the NIC and SATA controllers are passed through to TrueNAS VM
The VM has 150GB of memory and 8 vCPUs assigned to it by the host, the memory is reserved to the VM.

something else that I had forgotten to include in my initial post. the PCI SATA controller that the problem drives are connected to has been replaced. All SATA data cables have been replaced as well.
 

Davvo

MVP
Joined
Jul 12, 2022
Messages
3,222
Virtualization (without proper passthrough) and SATA controllers are both suspects.
 
Top