dm-brr(?) failure

Hi!

I have a big trouble with dm-bbr module. I have four drive in two 
mirror. I discovered data corruption in mysql databases which stored on 
this volmes and I found next messages in kern.log:

  Mar  9 12:00:32 server kernel: sd 0:0:2:0: SCSI error: return code = 
0x8000002
Mar  9 12:00:32 server kernel: sdc: Current: sense key=0x3
Mar  9 12:00:32 server kernel:     ASC=0x11 ASCQ=0x1
Mar  9 12:00:32 server kernel: Info fld=0x26af73c
Mar  9 12:00:32 server kernel: end_request: I/O error, dev sdc, sector 
40564540
Mar  9 12:00:32 server kernel: device-mapper: dm-bbr: device 8:32: I/O 
failure on sector 40564540. Scheduling for retry.
Mar  9 12:00:36 server kernel: sd 0:0:2:0: SCSI error: return code = 
0x8000002
Mar  9 12:00:36 server kernel: sdc: Current: sense key=0x3
Mar  9 12:00:36 server kernel:     ASC=0x11 ASCQ=0x1
Mar  9 12:00:36 server kernel: Info fld=0x26af73c
Mar  9 12:00:36 server kernel: end_request: I/O error, dev sdc, sector 
40564540
Mar  9 12:00:44 server kernel: sd 0:0:2:0: SCSI error: return code = 
0x8000002
Mar  9 12:00:44 server kernel: sdc: Current: sense key=0x3
Mar  9 12:00:44 server kernel:     ASC=0x11 ASCQ=0x1
Mar  9 12:00:44 server kernel: Info fld=0x26af73d
Mar  9 12:00:44 server kernel: end_request: I/O error, dev sdc, sector 
40564541
Mar  9 12:00:44 server kernel: device-mapper: dm-bbr: device 8:32: 
Trying to remap bad sector 40564541 to sector 139
Mar  9 12:00:49 server kernel: sd 0:0:2:0: SCSI error: return code = 
0x8000002
Mar  9 12:00:49 server kernel: sdc: Current: sense key=0x3
Mar  9 12:00:49 server kernel:     ASC=0x11 ASCQ=0x1
Mar  9 12:00:49 server kernel: Info fld=0x26af73f
Mar  9 12:00:49 server kernel: end_request: I/O error, dev sdc, sector 
40564543
Mar  9 12:00:49 server kernel: device-mapper: dm-bbr: device 8:32: 
Trying to remap bad sector 40564543 to sector 140

I got the next message from evmsn:

BBRseg: Warning: The kernel BBR segment manager is reporting that 4 bad 
sector(s) were discovered on segment sdc_bbr. Though replacement sectors 
are being used to remedy the problem, you would be well advised to take 
corrective actions by replacing the storage object.

But the raid1 driver can't deactivate the failed drive:

root@server:~# cat /proc/mdstat
Personalities : [raid1]
md255 : active raid1 dm-3[1] dm-2[0]
       35886464 blocks [2/2] [UU]

md0 : active raid1 dm-1[1] dm-0[0]
       35886464 blocks [2/2] [UU]


What can I do to prevent the data corruption?! Can I disable the bbr 
module without reinstalling the volumes?
This is an evms 2.5.0, but I discovered same behavior with 2.5.5 on sata 
drives. Thanks.

-- 
Gabor HALASZ <[email protected]>

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Evms-devel mailing list
[email protected]
To subscribe/unsubscribe, please visit:
https://lists.sourceforge.net/lists/listinfo/evms-devel