Archiving with vtls

mpeder

ADSM.ORG Member
Joined
Sep 17, 2007
Messages
6
Reaction score
0
Points
0
PREDATAR Control23

We are setting up an environment that is going to do an archive of a 20 TB flat file on a virtual tapes that belong to a cdl720. From what I am able to determine is this actually going to use 20TB's every day or will it just archive the new 500GBs that come in everyday? I am really hoping that it will only archive the new 500Gbs daily because they want this archived for seven years and that is going to eat up a lot of space.
 
PREDATAR Control23

I don't know the cdl720, what you are after is de-duplication rather than a vtl (or one that does both).
 
PREDATAR Control23

Deduplication doesn't come into play though until the new version which is tsm 6 on march 27th. Can dedupe be implemented on 5.5?
 
PREDATAR Control23

Archives are new complete copies each time. No versions, no incremental status.

Do you have any other media you can archive to? I would avoid putting long term archives on valuable CDL space and instead stick them on tape.
 
PREDATAR Control23

Nope the customer we are working for does not want anything on tape. They just want to use disk storage pools and vtls.
 
PREDATAR Control23

Deduplication doesn't come into play though until the new version which is tsm 6 on march 27th. Can dedupe be implemented on 5.5?

Yes but I'm not talking about tsm deduplication. You can get vtls that do it all for you, its hidden from tsm.

You are otherwise stuck I'm afraid. Even incremental backups would re-backup the entire file each time, unless you used sub-file backup but I'm not sure that would be a good idea for something of this size. You might need to look at changing the way the data is stored. 20tb is getting a bit ridiculous for a single file anyway.
 
Top