Bad HDD, HBA or Backplane?

[wookienz]

Hi all,

I am running openindiana VM on a ESXi box. The box has three M1015 HBA as pass through. Behind the hba's are 24 WD Red 3TB disks in 3 vdevs as RaidZ2

In /var/adm/messages i am getting alot of this:

May 10 08:27:03 oi scsi: [ID 365881 kern.info] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Log info 0x31080000 received for target 15.
May 10 08:27:03 oi scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
May 10 08:27:03 oi scsi: [ID 365881 kern.info] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Log info 0x31080000 received for target 15.
May 10 08:27:03 oi scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
May 10 08:27:03 oi scsi: [ID 107833 kern.notice] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Aborted_command!
May 10 08:27:03 oi scsi: [ID 243001 kern.warning] WARNING: /scsi_vhci (scsi_vhci0):
May 10 08:27:03 oi /scsi_vhci/disk@g50014ee003b6893d (sd30): Parity Error on path mpt_sas23/disk@w50014ee003b6893d,0
May 10 08:27:03 oi scsi: [ID 365881 kern.info] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Log info 0x31080000 received for target 15.
May 10 08:27:03 oi scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
May 10 08:27:03 oi scsi: [ID 365881 kern.info] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Log info 0x31080000 received for target 15.
May 10 08:27:03 oi scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
May 10 08:27:03 oi scsi: [ID 365881 kern.info] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Log info 0x31080000 received for target 15.
May 10 08:27:03 oi scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
May 10 08:27:03 oi scsi: [ID 365881 kern.info] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Log info 0x31080000 received for target 15.
May 10 08:27:03 oi scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
May 10 08:27:03 oi scsi: [ID 365881 kern.info] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Log info 0x31080000 received for target 15.
May 10 08:27:03 oi scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
May 10 08:27:03 oi scsi: [ID 107833 kern.notice] /pci@0,0/pci15ad,7a0@16/pci1000,3020@0 (mpt_sas6):
May 10 08:27:03 oi Aborted_command!
May 10 08:27:03 oi scsi: [ID 243001 kern.warning] WARNING: /scsi_vhci (scsi_vhci0):
May 10 08:27:03 oi /scsi_vhci/disk@g50014ee003b6893d (sd30): Parity Error on path mpt_sas23/disk@w50014ee003b6893d,0


a. is this the HBA, backplane or HDD?
b. how do i find the relevant port, hdd or hba based on the output?

Thanks.

 

[canta]

0x31080000 :
Type: 30000000h SAS
Origin: 01000000h PL
Code: 00080000h PL_LOGINFO_CODE_SATA_NCQ_FAIL_ALL_CMDS_AFTR_ERR

there is a simple python scripts ( or you can look on mpt2sas source to look lsi cryptic error definitions):
baruch/lsi_decode_loginfo · GitHub

"Log info 0x31080000 received for target 15" means NCQ command is failed on Target Drive 15.
you need to find which drive that reflect target 15.

On linux, the lsi error would point to /dev/sdX, which can be easy to cross-check with zpool status and /dev/disk/by-id.

someone has knowledge on open-solaris variance that would tell on how to know the location on "target 15".


good luck

 

[Lance Joseph]

0x31080000 :
Type: 30000000h SAS
Origin: 01000000h PL
Code: 00080000h PL_LOGINFO_CODE_SATA_NCQ_FAIL_ALL_CMDS_AFTR_ERR

there is a simple python scripts ( or you can look on mpt2sas source to look lsi cryptic error definitions):
baruch/lsi_decode_loginfo · GitHub

"Log info 0x31080000 received for target 15" means NCQ command is failed on Target Drive 15.
you need to find which drive that reflect target 15.

On linux, the lsi error would point to /dev/sdX, which can be easy to cross-check with zpool status and /dev/disk/by-id.

someone has knowledge on open-solaris variance that would tell on how to know the location on "target 15".


good luck

Thanks for the tip!
I'm still trying to figure out how the target # maps to a disk in illumos / openindiana.

 

[kroem]

Thanks for the tip!
I'm still trying to figure out how the target # maps to a disk in illumos / openindiana.

Did you manage to fix this issue? I'm getting the same logs...