Thanks Wanda,

you are completely right, I was checked this approach a day after my
question. For all other here, the correct way is:
1. Export archive node data (EXPORT NODE ... FILEDATA=ARCHIVE)
2. Delete archive data (by DELETE FILESPACE or by client GUI)
3. Change management class for archives (archive copy group) to point to new
tape pool
4. Import node data (IMPORT NODE)

Archive data will be imported to new pool regarding to set of management
class for given node.

Thanks to your help
Tom

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED]]On Behalf Of
Prather, Wanda
Sent: Thursday, November 08, 2001 5:39 PM
To: [EMAIL PROTECTED]
Subject: Re: Separate archive data


I've never done it, but for EXISTING data I THINK you could:

1) EXPORT the nodes and specify FILEDATA=ARCHIVE
That will put just the archive data on a tape.
2) Run DELETE FILESPACE for everyfile space and specify TYPE=ARCHIVE.
That will get rid of the archive data from the tape pool

That may or may not free a lot of tapes in the short term, but after RECLAIM
runs you should get the space back.

That takes a LOT of effort, and you no longer have a record of what was
archived since the DELETE FILESPACE takes it out of the data base.

Then set up separate pool for archives, and change your management classes
to point there.

Then I THINK you can even IMPORT the archives you have just exported (if you
do it BEFORE any new archives are created) and that will force the data into
the new pool...but I'm not positive about that, you'll have to test it.



-----Original Message-----
From: Toma9 Hrouda [mailto:[EMAIL PROTECTED]]
Sent: Thursday, November 08, 2001 5:26 AM
To: [EMAIL PROTECTED]
Subject: Separate archive data


Hi all,

is any way to separate archive data from backup data for some nodes (one
node in my case), when both are in one tape pool? (Mgmt class directs data
of both backup and archive copy groups to the same pool.)I need to remove
this archive data from tape library to offsite to clear space for backup
data (archives will be processed manualy one times in month).

Any idea to do it?

Thanks Tom

Reply via email to