Raid failure - drives or controller?

[Danilo Godec <danilo.godec@agenda.si>]

Hi,

I had two drive failure on a RAID5 in short time (unfortunately to short 
to rebuild on a spare disk). However - drives seem to work on a test 
machine and didn't report any errors. I also stuck them back into the 
orig. server (after rebooting) and they work now.

The first drive's errors were:

> Mar  6 05:15:19 san1 kernel: [10681162.473960] sd 4:0:3:0: [sde] 
> Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
> Mar  6 05:15:19 san1 kernel: [10681162.473965] sd 4:0:3:0: [sde] Sense 
> Key : Aborted Command [current]
> Mar  6 05:15:19 san1 kernel: [10681162.473969] sd 4:0:3:0: [sde] Add. 
> Sense: No additional sense information
> Mar  6 05:15:19 san1 kernel: [10681162.473973] sd 4:0:3:0: [sde] CDB: 
> Read(10): 28 00 07 af 38 3f 00 00 08 00
> Mar  6 05:15:19 san1 kernel: [10681162.473980] end_request: I/O error, 
> dev sde, sector 128923711
> Mar  6 05:17:53 san1 kernel: [10681316.885221] sd 4:0:3:0: [sde] 
> Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
> Mar  6 05:17:53 san1 kernel: [10681316.885225] sd 4:0:3:0: [sde] Sense 
> Key : Illegal Request [current]
> Mar  6 05:17:53 san1 kernel: [10681316.885229] sd 4:0:3:0: [sde] Add. 
> Sense: Logical block address out of range
> Mar  6 05:17:53 san1 kernel: [10681316.885234] sd 4:0:3:0: [sde] CDB: 
> Write(10): 2a 08 74 70 58 c7 00 00 08 00
> Mar  6 05:17:53 san1 kernel: [10681316.885242] end_request: I/O error, 
> dev sde, sector 1953519815
> Mar  6 05:17:53 san1 kernel: [10681316.885246] end_request: I/O error, 
> dev sde, sector 1953519815
> Mar  6 05:17:53 san1 kernel: [10681316.885252] raid5: Disk failure on 
> sde1, disabling device.
> Mar  6 05:20:27 san1 kernel: [10681470.600610] sd 4:0:3:0: [sde] 
> Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
> Mar  6 05:20:27 san1 kernel: [10681470.600615] sd 4:0:3:0: [sde] Sense 
> Key : Illegal Request [current]
> Mar  6 05:20:27 san1 kernel: [10681470.600619] sd 4:0:3:0: [sde] Add. 
> Sense: Logical block address out of range
> Mar  6 05:20:27 san1 kernel: [10681470.600624] sd 4:0:3:0: [sde] CDB: 
> Write(10): 2a 08 74 70 59 27 00 00 08 00
> Mar  6 05:20:27 san1 kernel: [10681470.600631] end_request: I/O error, 
> dev sde, sector 1953519911
> Mar  6 05:20:27 san1 kernel: [10681470.600636] end_request: I/O error, 
> dev sde, sector 1953519911
> Mar  6 05:20:28 san1 kernel: [10681471.664682]  disk 3, o:0, dev:sde1
> Mar  6 05:21:47 san1 kernel: [10681549.746852] sd 4:0:3:0: [sde] 
> Synchronizing SCSI cache
> Mar  6 05:21:47 san1 kernel: [10681549.746905] sd 4:0:3:0: [sde] 
> Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK

The second drive did this:

> Mar  7 02:31:37 san1 kernel: [10757598.197391] sd 4:0:5:0: [sdg] 
> Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
> Mar  7 02:31:37 san1 kernel: [10757598.197396] sd 4:0:5:0: [sdg] Sense 
> Key : Aborted Command [current]
> Mar  7 02:31:37 san1 kernel: [10757598.197400] sd 4:0:5:0: [sdg] Add. 
> Sense: No additional sense information
> Mar  7 02:31:37 san1 kernel: [10757598.197404] sd 4:0:5:0: [sdg] CDB: 
> Read(10): 28 00 07 12 05 9f 00 00 10 00
> Mar  7 02:31:37 san1 kernel: [10757598.197411] end_request: I/O error, 
> dev sdg, sector 118621599
> Mar  7 02:31:37 san1 kernel: [10757598.583990] raid5: Disk failure on 
> sdg1, disabling device.
> Mar  7 02:31:37 san1 kernel: [10757598.616232]  disk 5, o:0, dev:sdg1

Can anyone make some actual sense out of these sense messages?

Are these drives really / likely bad or is it more likely it was a 
controller failure?


    D.


-- 
Danilo Godec, sistemska podpora / system administration

Predlog! Obiscite prenovljeno spletno stran www.agenda.si

ODPRTA KODA IN LINUX
STORITVE : POSLOVNE RESITVE : UPRAVLJANJE IT : INFRASTRUKTURA IT : IZOBRAZEVANJE : PROGRAMSKA OPREMA

Visit our updated web page at www.agenda.si

OPEN SOURCE AND LINUX
SERVICES : BUSINESS SOLUTIONS : IT MANAGEMENT : IT INFRASTRUCTURE : TRAINING : SOFTWARE