Hi,
[please CC me on replies]
I've got a box running 2.6.10 (with the patch[0] needed to support the
Promise SATAII 150 TX4 controller).
This box has three software raid1 partitions mirrored on a SATA disk on
the Promise controller and a disk on the mainboard IDE controller (VIA
vt8235).
Within 4 days running the raid1, I got those three errors pasted below,
each marking the SATA-raidmember as faulty. After "raidhotremove" and
"raidhotadd" the SATA-raidmember syncs again fine and works at least a
day until it is marked as faulty again.
Any pointers where I could look at to resolve this problem?
The SATA drive is a new Seagate ST3250823AS.
Feb 6 04:49:04 mars kernel: ata4: status=0x51 { DriveReady SeekComplete
Error }
Feb 6 04:49:04 mars kernel: SCSI error : <3 0 0 0> return code = 0x8000002
Feb 6 04:49:04 mars kernel: FMK Current sda: sense = 70 88
Feb 6 04:49:04 mars kernel: ASC=40 ASCQ=c0
Feb 6 04:49:04 mars kernel: end_request: I/O error, dev sda, sector
311900096
Feb 6 04:49:04 mars kernel: raid1: Disk failure on sda2, disabling device.
Feb 6 04:49:04 mars kernel: ^IOperation continuing on 1 devices
Feb 6 04:49:04 mars kernel: raid1: sda2: rescheduling sector 302518136
Feb 6 04:49:04 mars kernel: RAID1 conf printout:
Feb 6 04:49:04 mars kernel: --- wd:1 rd:2
Feb 6 04:49:04 mars kernel: disk 0, wo:1, o:0, dev:sda2
Feb 6 04:49:04 mars kernel: disk 1, wo:0, o:1, dev:hda2
Feb 6 04:49:04 mars kernel: RAID1 conf printout:
Feb 6 04:49:05 mars kernel: --- wd:1 rd:2
Feb 6 04:49:05 mars kernel: disk 1, wo:0, o:1, dev:hda2
Feb 6 04:49:05 mars kernel: raid1: hda2: redirecting sector 302518136
to another mirror
Feb 7 06:25:18 mars kernel: ata4: status=0x51 { DriveReady SeekComplete
Error }
Feb 7 06:25:18 mars kernel: SCSI error : <3 0 0 0> return code = 0x8000002
Feb 7 06:25:18 mars kernel: FMK Current sda: sense = 70 88
Feb 7 06:25:18 mars kernel: ASC=40 ASCQ=c0
Feb 7 06:25:18 mars kernel: end_request: I/O error, dev sda, sector
364654755
Feb 7 06:25:18 mars kernel: raid1: Disk failure on sda5, disabling device.
Feb 7 06:25:18 mars kernel: ^IOperation continuing on 1 devices
Feb 7 06:25:18 mars kernel: raid1: sda5: rescheduling sector 3706272
Feb 7 06:25:18 mars kernel: RAID1 conf printout:
Feb 7 06:25:18 mars kernel: --- wd:1 rd:2
Feb 7 06:25:18 mars kernel: disk 0, wo:1, o:0, dev:sda5
Feb 7 06:25:18 mars kernel: disk 1, wo:0, o:1, dev:hda5
Feb 7 06:25:18 mars kernel: RAID1 conf printout:
Feb 7 06:25:18 mars kernel: --- wd:1 rd:2
Feb 7 06:25:18 mars kernel: disk 1, wo:0, o:1, dev:hda5
Feb 7 06:25:18 mars kernel: raid1: hda5: redirecting sector 3706272 to
another mirror
Feb 9 06:25:02 mars kernel: ata4: status=0x51 { DriveReady SeekComplete
Error }
Feb 9 06:25:02 mars kernel: SCSI error : <3 0 0 0> return code = 0x8000002
Feb 9 06:25:02 mars kernel: FMK Current sda: sense = 70 88
Feb 9 06:25:02 mars kernel: ASC=40 ASCQ=c0
Feb 9 06:25:02 mars kernel: end_request: I/O error, dev sda, sector
310063304
Feb 9 06:25:02 mars kernel: raid1: Disk failure on sda2, disabling device.
Feb 9 06:25:02 mars kernel: ^IOperation continuing on 1 devices
Feb 9 06:25:02 mars kernel: raid1: sda2: rescheduling sector 300681344
Feb 9 06:25:02 mars kernel: RAID1 conf printout:
Feb 9 06:25:02 mars kernel: --- wd:1 rd:2
Feb 9 06:25:02 mars kernel: disk 0, wo:1, o:0, dev:sda2
Feb 9 06:25:02 mars kernel: disk 1, wo:0, o:1, dev:hda2
Feb 9 06:25:02 mars kernel: RAID1 conf printout:
Feb 9 06:25:02 mars kernel: --- wd:1 rd:2
Feb 9 06:25:02 mars kernel: disk 1, wo:0, o:1, dev:hda2
Feb 9 06:25:02 mars kernel: raid1: hda2: redirecting sector 300681344
to another mirror
[0] http://marc.theaimsgroup.com/?l=linux-ide&m=110426005503319&q=raw
Best regards,
-jr
Johannes Resch wrote:
> Hi,
>
> [please CC me on replies]
>
> I've got a box running 2.6.10 (with the patch[0] needed to support the
> Promise SATAII 150 TX4 controller).
> This box has three software raid1 partitions mirrored on a SATA disk on
> the Promise controller and a disk on the mainboard IDE controller (VIA
> vt8235).
>
> Within 4 days running the raid1, I got those three errors pasted below,
> each marking the SATA-raidmember as faulty. After "raidhotremove" and
> "raidhotadd" the SATA-raidmember syncs again fine and works at least a
> day until it is marked as faulty again.
>
> Any pointers where I could look at to resolve this problem?
> The SATA drive is a new Seagate ST3250823AS.
I would change out your cables, and also make sure you are running
2.6.11-rc3-bk-latest, which includes all the SATAII patches and other fixes.
JEff
Jeff Garzik wrote::
>Johannes Resch wrote:
>> Hi,
>>
>> [please CC me on replies]
>>
>> I've got a box running 2.6.10 (with the patch[0] needed to support the
>> Promise SATAII 150 TX4 controller).
>> This box has three software raid1 partitions mirrored on a SATA disk on
>> the Promise controller and a disk on the mainboard IDE controller (VIA
>> vt8235).
>>
>> Within 4 days running the raid1, I got those three errors pasted below,
>> each marking the SATA-raidmember as faulty. After "raidhotremove" and
>> "raidhotadd" the SATA-raidmember syncs again fine and works at least a
>> day until it is marked as faulty again.
>>
>> Any pointers where I could look at to resolve this problem?
>> The SATA drive is a new Seagate ST3250823AS.
>I would change out your cables, and also make sure you are running
>2.6.11-rc3-bk-latest, which includes all the SATAII patches and other fixes.
I don't believe it has anything to do with cabling. 2.6.10-ac9 introduced
some sata patches. I didn't check -ac9 and -ac10, but -ac11 and -ac12 are
not usable on my box with exactly the same symptoms.
-jo
--
-rw-r--r-- 1 jo users 63 2005-02-12 18:43 /home/jo/.signature