Disk read errors - bad HBA, bad drive cage, bad drives?

[DayspringGaming]

I've been fighting this for a week or two now. A drive will just disable from the array. So I'll make sure everything's off the drive, run an extended SMART test but not see major errors. This has now happened 3 times. Need to figure out where my problem is here so I can fix it. The 3 drives here are all shucked WD Whites. HBA is an LSI 9208, and they're sitting in a Rosewill hot swap 4 drive cage.

 Attaching syslog and the SMART drive readouts for the 3 drives in question.

syslog.txt WDC_WD80EMAZ_20210129-1033.txt WDC_WD100EMAZ_20210129-1033.txt WDC_WD120EDAZ_20210129-1033.txt

[JorgeB]

Next time please post the full diagnostics, but simultaneous errors on multiple disks

 

Jan 29 10:21:58 PapandriaServer kernel: md: disk6 read error, sector=1332952
Jan 29 10:21:58 PapandriaServer kernel: md: disk6 read error, sector=1332960
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#319 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#319 Sense Key : 0x2 [current]
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#319 ASC=0x4 ASCQ=0x0
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#319 CDB: opcode=0x8a 8a 00 00 00 00 00 00 00 00 40 00 00 00 08 00 00
Jan 29 10:21:58 PapandriaServer kernel: print_req_error: I/O error, dev sdb, sector 64
Jan 29 10:21:58 PapandriaServer kernel: md: disk6 write error, sector=0
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#320 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#320 Sense Key : 0x2 [current]
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#320 ASC=0x4 ASCQ=0x0
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:0:0: [sdb] tag#320 CDB: opcode=0x8a 8a 00 00 00 00 00 00 00 00 58 00 00 00 10 00 00
Jan 29 10:21:58 PapandriaServer kernel: print_req_error: I/O error, dev sdb, sector 88
Jan 29 10:21:58 PapandriaServer kernel: md: disk6 write error, sector=24
Jan 29 10:21:58 PapandriaServer kernel: md: disk6 write error, sector=32
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:2:0: [sdd] tag#323 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:2:0: [sdd] tag#323 Sense Key : 0x2 [current]
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:2:0: [sdd] tag#323 ASC=0x4 ASCQ=0x0
Jan 29 10:21:58 PapandriaServer kernel: sd 6:0:2:0: [sdd] tag#323 CDB: opcode=0x88 88 00 00 00 00 00 00 00 00 c0 00 00 00 20 00 00
Jan 29 10:21:58 PapandriaServer kernel: print_req_error: I/O error, dev sdd, sector 192
Jan 29 10:21:58 PapandriaServer kernel: md: disk7 read error, sector=128
Jan 29 10:21:58 PapandriaServer kernel: md: disk7 read error, sector=136
Jan 29 10:21:58 PapandriaServer kernel: md: disk7 read error, sector=144
Jan 29 10:21:58 PapandriaServer kernel: md: disk7 read error, sector=152

 

suggest a controller, power/connection problem.

[DayspringGaming]

@JorgeB

Attached below.

diagnostics_files.zip

[JorgeB]

I don't need them now, was just saying for next time, since the syslog is included as well as much more info, for now see if the failing disks share something other than the HBA and PSU, like a miniSAS cable or power splitter, if not try another PSU or HBA.

[DayspringGaming]

They share all of the following:

Drive cage (handles power & SATA connections to the drive)

HBA

SFF 8087 Cable

I'm looking into a new HBA. Next after that is a PSU upgrade (650W right now) and instead of using the rosewill cage, do a direct connection to the drives.