I have a script that I use for labeling luns. So if I add 100 luns from my SAN
I run the script and it runs devfsadm or cfgadm where appropriate and then gets
a list of unlabeled luns via the format command and then iterates over the list
and issues the following command...
format -f /tmp/label -d cxtxdx
:/tmp/label
type 0
label
On my Solaris 10 Update 3 systems each label operation takes about 30 seconds.
When I truss the command I see many system calls for every path to every lun.
Some calls are issued for every slice of every lun on every path.
Here is an snippet of output grepped from the truss output...
6327: 0.4289 0.0001 0.0000 chdir("/dev/rdsk") = 0
6327: 0.4308 0.0002 0.0000 open("/dev/rdsk/c8t81d187s2", O_RDWR|O_NDELAY) = 4
6327: 0.4333 0.0001 0.0000 getcwd("/dev/rdsk", 1024) = 0
6327: 0.4334 0.0001 0.0000 chdir("/dev/rdsk") = 0
6327: 0.4335 0.0001 0.0000 stat("/dev/rdsk/c8t81d187s2", 0xFFBFE088) = 0
6327: 0.4336 0.0001 0.0000 lstat("/dev/rdsk/c8t81d187s2", 0xFFBFE088) = 0
6327: 0.4337 0.0001 0.0000 readlink("/dev/rdsk/c8t81d187s2",
"../../devices/[EMAIL PROTECTED],700000/[EMAIL PROTECTED],
1/[EMAIL PROTECTED],bb:c,raw", 1024) = 51
Is it expected behavior for format to perform so many operations on all luns
when the command is evoked with the "-d" switch?
Thanks,
Vic
This message posted from opensolaris.org
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss
