This post is more than 5 years old
12 Posts
0
695
Solaris 8 VCS Cluster Nodes and Failed Disk in INQ
Hello all
I have two Solaris 8 VCS clusters attached to a cx700. Recently, on each cluster, I removed access to some unused LUNs. These LUNs still appear in INQ as follows (please ignore the smileys from the copy/paste):
$ sudo /opt/sanadmin/bin/inq -nodots |grep FAILED
/dev/rdsk/c3t3d3s2 GC : 226 3000000 : FAILED
/dev/rdsk/c3t3d6s2 GC : 226 6000000 : FAILED
/dev/rdsk/c3t3d7s2 GC : 226 7000000 : FAILED
/dev/rdsk/c4t4d3s2 GC : 226 3000000 : FAILED
/dev/rdsk/c4t4d6s2 GC : 226 6000000 : FAILED
/dev/rdsk/c4t4d7s2 GC : 226 7000000 : FAILED
Running devfsadm doesn't change anything. Is this a hung scsi reservation? I am worried that the only way to fix this will be to reboot, so I thought I would gather any thoughts/tips I can.
Any help is greatly appriciated!
Conor
341 Posts
0
January 6th, 2010 13:00
OK, to me this looks like only a reconfig reboot will clear these 'ghost' devices
Conor
341 Posts
0
January 6th, 2010 00:00
Do the same devices show in the output of 'format'? Or is it just 'inq'?
Have you tried running
devfsadm -C
To cleanup the device tree (remove dead device paths?)
Conor
ksmith123
12 Posts
0
January 6th, 2010 08:00
The disk is visible in both format and INQ. devfsadm -C doesn't clear up the issue.
ksmith123
12 Posts
0
January 7th, 2010 07:00
That's what I was afraid of - bummer!
Thanks for the input.