[SunHELP] SCSI transport error

[Leslie V Brigance]

What type SENAs & how connected?

Les

Subject:  [SunHELP] SCSI transport error

Hi all,

I am on a client site and having problems on a E4000 running Solaris 2.6,
with 6 CPUs, 4GB memory and 4 SENAs. Using Veritas Volume Manager, all
disks on RAID5s. That is their NFS server.
/var/adm/messages shows the following error:

Nov  9 11:23:01 ang056 unix: WARNING:
/sbus at b,0/SUNW,socal at 0,0/sf at 0,0/ssd at w2200002037c98530,0 (ssd238):^M^M
Nov  9 11:23:01 ang056 unix:    SCSI transport failed: reason 'timeout':
giving up^M^M
Nov  9 11:36:56 ang056 unix: NOTICE: vxdmp: Path failure on 118/0x774
Nov  9 11:36:56 ang056 unix: NOTICE: vxvm:vxdmp: disabled path 118/0x770
belonging to the dmpnode 158/0x2a8^M
Nov  9 11:37:00 ang056 unix: NOTICE: vxvm:vxdmp: enabled path 118/0x770
belonging to the dmpnode 158/0x2a8^M
Nov  9 11:38:01 ang056 unix: WARNING:
/sbus at b,0/SUNW,socal at 0,0/sf at 1,0/ssd at w2100002037c98530,0 (ssd259):^M^M
Nov  9 11:38:01 ang056 unix:    SCSI transport failed: reason 'timeout':
retrying command^M^M
Nov  9 11:41:16 ang056 unix: WARNING:
/sbus at b,0/SUNW,socal at 0,0/sf at 1,0/ssd at w2100002037c98530,0 (ssd259):^M^M
Nov  9 11:41:16 ang056 unix:    SCSI transport failed: reason 'timeout':
giving up^M^M

The SENA that contains this disk does not show any error/problem. There was
the same error for a different disk ( c7t38d0, same filesystem, same SENA)
early this morning and i put it offline, error gone for that disk and
started again with the one above (c7t58d0). I cannot offline this one
without causing problems to the filesystem.
I noticed that luxadm display would freeze, for that particular SENA, while
the system was "giving up".
Can someone help me on this?

Many thanks,
Joe