i have a host - hpux 11.0 , cx 3-80 3 LUNS - lun 1 , 2 , 3 (already have file systems and lots of free space on VG's) i use clarion snaps .
last night , i need to recover something from snap , so i created 4 files systems using above LUNS . i mounted the snap luns on the same host and started copying from snap luns (which i mounted ) to newly created file systems . After copying for about 2-3 hours , i started seeing some i/o error on the system . I checked the navi and founf that my session associated with the LUN are gone . I just the navi log and found out that the Snap LUN got full (100%) and probably clariion removed them due to some integrity issues.
I think i figure out that it is not practice to mount the snap on the same host but like to know your opinion and some root cause.
i can mount the snap lun to the different a different host . but i still need to perform copy and this will again increase the writes and grow the snap reserved luns . Is there a way i can stop the changes from happening to my Snap LUNS while iam doing the copy . This way i don't have to worry about the filling my reserver snap luns .
if you must mount snapshots/clones on the same host as the source volume, make sure to use vgchgid before you import the disk. You can use this solution, even though it talks about BCVs, you need to pefrom the steps on your target devices.