[TriLUG] SATA/RAID issues on SiS 180/181 with WD 2000JS 200GB

[Brian Daniels]

I've been fighting with a SATA RAID issue all week.  Anyone seen this problem 
before?

I have been having command timeouts and CRC issues with the following setup:

AMD64 MB with SiS 180/181 SATA controller.
Two WD drives: Vendor: ATA       Model: WDC WD2000JS-00P  Rev: 21.0
Drives are in a software RAID1 array.
Centos 4.4 x86_64 running kernel 2.6.9-42.0.2.EL

The drives generally work fine.  But under heavy load (like a RAID mirror 
rebuild or large SQL dump) I get errors like these:

Sep  5 13:39:58 rt3 kernel: ata1: translated ATA stat/err 0x51/84 to SCSI 
SK/ASC/ASCQ 0xb/47/00
Sep  5 13:39:58 rt3 kernel: ata1: status=0x51 { DriveReady SeekComplete Error }
Sep  5 13:39:58 rt3 kernel: ata1: error=0x84 { DriveStatusError BadCRC }

eventually they will cause more serious problems:

Sep  7 21:38:20 rt3 kernel: ata1: command 0xc8 timeout, stat 0x51 host_stat 0x1
Sep  7 21:38:20 rt3 kernel: ata1: translated ATA stat/err 0x51/84 to SCSI 
SK/ASC/ASCQ 0xb/47/00
Sep  7 21:38:20 rt3 kernel: ata1: status=0x51 { DriveReady SeekComplete Error }
Sep  7 21:38:20 rt3 kernel: ata1: error=0x84 { DriveStatusError BadCRC }
Sep  7 21:38:20 rt3 kernel: SCSI error : <0 0 0 0> return code = 0x8000002
Sep  7 21:38:20 rt3 kernel: Info fld=0x484226c, Current sda: sense key Aborted 
Command
Sep  7 21:38:20 rt3 kernel: Additional sense: Scsi parity error
Sep  7 21:38:20 rt3 kernel: end_request: I/O error, dev sda, sector 75768173
Sep  7 21:38:20 rt3 kernel: raid1: Disk failure on sda2, disabling device.
Sep  7 21:38:20 rt3 kernel:     Operation continuing on 1 devices
Sep  7 21:38:20 rt3 kernel: raid1: sda2: rescheduling sector 75559328
Sep  7 21:38:20 rt3 kernel: raid1: sdb2: redirecting sector 75559328 to another 
mirror

The error usually occurs on ata1, but sometimes ata2 is blamed.

I have tried:
New SATA cables.
Returning drive and exchanging for new identical unit.
New power supply.
Setting jumper on drive to force SATA 150 MB/s mode.
Moving drives to another system with the same model motherboard.

None of these helped - the problem would eventually reoccur.

Strangely, I have two older models of this drive:
Vendor: ATA       Model: WDC WD2000JD-22H  Rev: 08.0
that work perfectly in the same systems, in the same arrangement.


Thanks,
Brian

-- 
Be it thy course to busy giddy minds
With foreign quarrels; that action, hence borne out,
May waste the memory of the former days.
			Henry IV, Part 2
	

Brian Daniels                  bitmage at bellsouth.net
      http://www.eviloverlord.net