SLES server connected to FC storage (SCSI error while I/O)
 

I have SLES10SP4(x86_64) installed on my server connected with SAS storage attached externally via FC cables.. I have configured DM-Multipath on my server. My storage has redundant controllers so that when I switch off one controller, the other can be used.
I start continuous I/O from my server on to the storage and then I reboot one of my controller. Then the I/O on my server breaks inspite of the fact that the other controller is active. I get the following messages in the /var/log/messages. I tried analyzing the SCSI error code., but could not understand them.
Please help.

Jul 18 16:38:37 sles10sp4-64bit kernel: JBD: barrier-based sync failed on dm-5 - disabling barriers
Jul 18 16:39:35 sles10sp4-64bit kernel: lpfc 0000:06:00.0: 2:1305 Link Down Event x6 received Data: x6 x20 x80000 x0 x0
Jul 18 16:40:05 sles10sp4-64bit kernel: rport-6:0-0: blocked FC remote port time out: saving binding
Jul 18 16:40:05 sles10sp4-64bit kernel: sd 6:0:0:11: SCSI error: return code = 0x00010000
Jul 18 16:40:05 sles10sp4-64bit kernel: end_request: I/O error, dev sdq, sector 0
Jul 18 16:40:05 sles10sp4-64bit kernel: sd 6:0:0:14: SCSI error: return code = 0x00010000
Jul 18 16:40:05 sles10sp4-64bit kernel: end_request: I/O error, dev sdr, sector 0
Jul 18 16:40:05 sles10sp4-64bit kernel: sd 6:0:0:16: SCSI error: return code = 0x00010000
Jul 18 16:40:05 sles10sp4-64bit kernel: end_request: I/O error, dev sds, sector 34656
Jul 18 16:40:05 sles10sp4-64bit kernel: device-mapper: multipath: Failing path 65:32.

Looks like you may need to adjust your multipath configuration to allow for proper failover. What type of backend SAN are you working with? Also what does your multipath.conf look like?

Make sure to use Code tags.

The multipath.conf file looks like:
defaults
{
user_friendly_names yes
failback immediate
prio_callout "/sbin/mpath_prio_alua /dev/%n"
}

I am using FC SAN storage.