Heavy harddisk or controller errors

malo_umoran · 07-26-2007, 07:52 AM

Hi,

I have since yesterday some have problems with my computer. It started to boot and than it showed the following messages:

Code:

Jul 26 10:29:58 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x100000 action 0x2
Jul 26 10:29:58 jingle kernel: ata3.00: (BMDMA stat 0x45)
Jul 26 10:29:58 jingle kernel: ata3.00: cmd c8/00:00:04:81:1f/00:00:00:00:00/e0 tag 0 cdb 0x0 data 131072 in
Jul 26 10:29:58 jingle kernel:          res 51/84:8f:75:81:1f/84:02:02:00:00/e0 Emask 0x10 (ATA bus error)
Jul 26 10:29:58 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x180000 action 0x2
Jul 26 10:29:58 jingle kernel: ata3.00: (BMDMA stat 0x45)
Jul 26 10:29:58 jingle kernel: ata3.00: cmd c8/00:80:84:ff:01/00:00:00:00:00/e0 tag 0 cdb 0x0 data 65536 in
Jul 26 10:29:58 jingle kernel:          res 51/84:1f:e5:ff:01/84:02:02:00:00/e0 Emask 0x10 (ATA bus error)
Jul 26 10:29:58 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x100000 action 0x2 frozen
Jul 26 10:29:58 jingle kernel: ata3.00: cmd ca/00:20:e4:37:01/00:00:00:00:00/e0 tag 0 cdb 0x0 data 16384 out
Jul 26 10:29:58 jingle kernel:          res 40/00:1f:e5:ff:01/84:02:02:00:00/e0 Emask 0x4 (timeout)
Jul 26 10:29:58 jingle kernel: ata3: port is slow to respond, please be patient (Status 0xd0)
Jul 26 10:29:58 jingle kernel: ata3: port failed to respond (30 secs, Status 0xd0)
Jul 26 10:29:58 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x100000 action 0x2 frozen
Jul 26 10:29:58 jingle kernel: ata3.00: cmd ca/00:20:e4:37:01/00:00:00:00:00/e0 tag 0 cdb 0x0 data 16384 out
Jul 26 10:29:58 jingle kernel:          res 40/00:1f:e5:ff:01/84:02:02:00:00/e0 Emask 0x4 (timeout)
Jul 26 10:29:58 jingle kernel: ata3: port is slow to respond, please be patient (Status 0xd0)
Jul 26 10:29:58 jingle kernel: ata3: port failed to respond (30 secs, Status 0xd0)
Jul 26 10:29:58 jingle kernel: ata3: port is slow to respond, please be patient (Status 0xd0)
Jul 26 10:29:58 jingle kernel: ata3: port failed to respond (30 secs, Status 0xd0)
Jul 26 10:29:58 jingle kernel: ATA: abnormal status 0xD0 on port 0x000109f7
Jul 26 10:29:58 jingle last message repeated 6 times
Jul 26 10:29:58 jingle kernel: ata3.00: qc timeout (cmd 0xec)
Jul 26 10:29:58 jingle kernel: ata3.00: failed to IDENTIFY (I/O error, err_mask=0x4)
Jul 26 10:29:58 jingle kernel: ata3.00: revalidation failed (errno=-5)
Jul 26 10:29:58 jingle kernel: ata3: failed to recover some devices, retrying in 5 secs
Jul 26 10:29:58 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x180000 action 0x2
Jul 26 10:29:58 jingle kernel: ata3.00: (BMDMA stat 0x44)
Jul 26 10:29:58 jingle kernel: ata3.00: cmd c8/00:20:54:0a:03/00:00:00:00:00/e0 tag 0 cdb 0x0 data 16384 in
Jul 26 10:29:58 pornost4r kernel:          res 51/84:00:73:0a:03/84:02:02:00:00/e0 Emask 0x10 (ATA bus error)
Jul 26 10:30:00 pornost4r kernel: Module len 1548288 truncated
Jul 26 10:30:00 pornost4r kernel: Module len 1548288 truncated
Jul 26 10:30:01 pornost4r kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x380000 action 0x2

It took around 5-10 minutes to boot, but everything looked OK. I made a rsync-backup right after that and I saw in syslog some similar messages while rsync was running.

I recompiled new kernel (2.6.22.1), but it did not help much. Boot was OK, but I saw the messages again, and disk was terribly slow from time to time. I saw a lot of similar messages in syslog during these "slow periods":

Code:

Jul 26 13:07:50 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x100000 action 0x2 frozen
Jul 26 13:07:50 jingle kernel: ata3.00: cmd 35/00:00:04:5a:d3/00:04:00:00:00/e0 tag 0 cdb 0x0 data 524288 out
Jul 26 13:07:50 jingle kernel:          res 40/00:00:8b:fb:34/00:00:00:00:00/e0 Emask 0x4 (timeout)
Jul 26 13:07:55 jingle kernel: ata3: port is slow to respond, please be patient (Status 0xd0)
Jul 26 13:08:00 jingle kernel: ata3: device not ready (errno=-16), forcing hardreset
Jul 26 13:19:48 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x180000 action 0x2
Jul 26 13:19:48 jingle kernel: ata3.00: (BMDMA stat 0x45)
Jul 26 13:19:48 jingle kernel: ata3.00: cmd c8/00:b0:14:0d:f5/00:00:00:00:00/e0 tag 0 cdb 0x0 data 90112 in
Jul 26 13:19:48 jingle kernel:          res 51/84:1f:a5:0d:f5/00:00:00:00:00/e0 Emask 0x10 (ATA bus error)
Jul 26 13:19:49 jingle kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x100000 action 0x2
Jul 26 13:19:49 jingle kernel: ata3.00: (BMDMA stat 0x45)

What do you think, is this a hard disk problem or a controller problem? I have a motherboard (Asus M2N32-SLI) with 6 SATA slots (nforce590) and 5 hard disks. 3 disks are in raid5 and I see no problems with these disks. As I can see, only /dev/sda is making problems, so it should be hard disk and not a controller.

But, on the other side, syslog says:

Code:

Jul 26 13:07:55 jingle kernel: ata3: port is slow to respond, please be patient (Status 0xd0)
Jul 26 13:08:00 jingle kernel: ata3: device not ready (errno=-16), forcing hardreset

and that sounds for me like a libata / controller error.

I found a few partly usefull posts on the net, but nothing which would solve this problem.

Any ideas, advices etc?
Thanks a lot

M.

Hern_28 · 07-26-2007, 08:45 AM

From the looks of the log and due to the fact other drives on that controller are working properly I would download test utilities from the hard drive manufacturer and run checks on the drive. If you wanted to do some extra testing also could temporarily swap the drive cable, and if that doesn't make for a performance fix could try connecting the drive to a different port to rule a channel problem. The errors look similar to the ones I got when my sata drive overheated though ( 1 month after warrenty expiration

lol )

malo_umoran · 07-26-2007, 08:56 AM

I was also thinking about that ...
1. check cable
2. check sata slot
3. check disk (utilities)
4. hmmm ... replace motherboard

onebuck · 07-26-2007, 01:11 PM

Hi,

I would first make sure you do have valid backups. Then I would get the diagnostics for the drives. I would not rule out memory either.

malo_umoran · 07-27-2007, 07:32 PM

Quote:

Originally Posted by onebuck

I would first make sure you do have valid backups. Then I would get the diagnostics for the drives. I would not rule out memory either.

As I already told in the first message:

Quote:

It took around 5-10 minutes to boot, but everything looked OK. I made a rsync-backup right after that and I saw in syslog some similar messages while rsync was running.

I have daily rsync backup and I was not worried. I let badblocks and smartctl to check the disk, but nothing was found.

I found the reason for the problems after opening the case. SATA cable was the guilty subject.

I wanted to check cable before I replace the disk and I was able to pull it out very easy. Too easy. I replaced it with another one, and everything is OK since.

Thanks.

M.

onebuck · 07-28-2007, 07:51 AM

Hi,

Thanks for the feedback and closing the post!