In your script check the number of copies in the mminfo output you use to generate your clone/stage list - if copies=1 then it is uncloned, if copies>1 then it has been cloned. You also may wish to keep the backup on disk for several days to facilitate recoveries, make your tape clone copy immediately for offsite backups, then stage the data to onsite tape after several days to free up disk space.
The clone will be identical to the original, that is why it is a clone?
* write to disk 1 (and possibly disk 2) * clone disk 1 to pool 1 * stage disk 1 to pool 2
The reasoning would be that virtual devices tend to be cheaper and thus you may have more available so can match 1 to 1 devices with physical devices, plus there would be some benefits regarding tape handling.
You can't be sure as that depends on staging policy and available space on disk in the first place. What is always a good practice is to have script which checks disk and tape; if it was staged before cloned that script checking up tapes will pick it up.
Well my scripting is Solaris and it may not fit what you wish to do anyway (plus it will be over 700 lines long when I get the load balancing function in it so it will be pretty confusing for you).
DavidHampson
2 Intern
•
1080 Posts
449
0
Posted July 1st, 2009 01:00
The clone will be identical to the original, that is why it is a clone?