Joe, > From what I am reading I am able to perform this operation at the TSM > server and it directs the output to scratch tapes. The GREAT thing about > this is it does NOT touch the serves to retransmit all the data to the > archive storage pool, and then to tape (which is alot of traffic!). The > Backups sets would use more tapes (kind of like collocation turned on > right?!). So after the daily incremental running at night, I will be > able to create the backup sets ( for a monthly archive) using the > current data just backed up from that node. (right?!) Right! The backupset-"archive" gets created without having to retransmit any data from the node to the server.
> One question is when doing a backup set, does it also save all the > inactive files? or just the active ones? Like an archive, a backup set only contains active data. > Is there a way to direct all the backup sets to another copy pool to > utilize tape consumption and to keep the tapes separate from the other > copypools for easier search? (just a thought - but this would be a lot > of extra work. Backupsets aren't written to a storage pool. You just define the device class to be used and tsm takes the tapes from the scratch list of your checked in library volumes. On a "q libvol" you can see the tapes like this: IBM3494B C00246 Private Data IBM3494B C00247 Private BackupSet IBM3494B C00248 Private Data IBM3494B C00249 Scratch IBM3494B C00251 Private DbBackup One advantage of a backupset is that the tape containing the backupset can get used on another tsm server to restore the client. This is not possible with an archive. Best regards, Michael -- Michael Bartl mailto:[EMAIL PROTECTED] Office of Technology, IT Germany/Austria Tel: +49-89-92699-806 Cable & Wireless Deutschland GmbH. Fax: +49-89-92699-302 Landsberger Str. 155, D-80687 Muenchen http://www.cw.com/de
