6.9.2 SSD in cache pool failing?

[fysmd]

Hi, I am running 6.9.2 and noticed errors in syslog this morning which I think suggest one of mt cache pool drives is misbehaving but there are no errors displayed anywhere, nor any note on that drive - are these just info or should I change this drive?

 

 

Dec 30 07:44:06 Server kernel: sd 7:0:8:0: attempting task abort!scmd(0x0000000094d60fd2), outstanding for 30352 ms & timeout 30000 ms
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: [sdp] tag#4543 CDB: opcode=0x28 28 00 03 20 d8 00 00 00 08 00
Dec 30 07:44:06 Server kernel: scsi target7:0:8: handle(0x0021), sas_address(0x300062b202f04009), phy(9)
Dec 30 07:44:06 Server kernel: scsi target7:0:8: enclosure logical id(0x500062b202f04000), slot(18) 
Dec 30 07:44:06 Server kernel: scsi target7:0:8: enclosure level(0x0000), connector name(     )
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: task abort: SUCCESS scmd(0x0000000094d60fd2)
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: [sdp] tag#4543 UNKNOWN(0x2003) Result: hostbyte=0x03 driverbyte=0x00 cmd_age=30s
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: [sdp] tag#4543 CDB: opcode=0x28 28 00 03 20 d8 00 00 00 08 00
Dec 30 07:44:06 Server kernel: blk_update_request: I/O error, dev sdp, sector 52484096 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: attempting task abort!scmd(0x00000000d67ffc34), outstanding for 30382 ms & timeout 30000 ms
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: [sdp] tag#4541 CDB: opcode=0x28 28 00 03 20 56 08 00 00 08 00
Dec 30 07:44:06 Server kernel: scsi target7:0:8: handle(0x0021), sas_address(0x300062b202f04009), phy(9)
Dec 30 07:44:06 Server kernel: scsi target7:0:8: enclosure logical id(0x500062b202f04000), slot(18) 
Dec 30 07:44:06 Server kernel: scsi target7:0:8: enclosure level(0x0000), connector name(     )
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: No reference found at driver, assuming scmd(0x00000000d67ffc34) might have completed
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: task abort: SUCCESS scmd(0x00000000d67ffc34)
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: [sdp] tag#4541 UNKNOWN(0x2003) Result: hostbyte=0x03 driverbyte=0x00 cmd_age=30s
Dec 30 07:44:06 Server kernel: sd 7:0:8:0: [sdp] tag#4541 CDB: opcode=0x28 28 00 03 20 56 08 00 00 08 00
Dec 30 07:44:06 Server kernel: blk_update_request: I/O error, dev sdp, sector 52450824 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0

 

complete diag attached.

server-diagnostics-20221230-0909.zip

[JorgeB]

Looks more like a power/connection problem, also by having the SSDs on the HBA you'll lose trim, suggest swapping them with disks using the onboard SATA ports.

[fysmd]

Ahh - OK,

I did read about losing trim but I thought that was related to the devices which allow four drive to user one host port.

Will make some changes,  Thanks!

[fysmd]

Can you explain how you determine power issues from logs - what to look for for me please?

I'd like to me able to identify such myself.

 

Edited December 31, 2022 by fysmd

[JorgeB]

The type of error, when it's a media error the error it's usually different, also this:

 

Dec 30 07:44:07 Server kernel: sd 7:0:8:0: Power-on or device reset occurred