Conference decwet::networker

508.0. "Testing assumptions/learning about cloning" by SANITY::LEMONS (And we thank you for your support.) Wed Mar 19 1997 21:16

    Hi
    
    We're creating a pool of tapes that will have a one year retention and
    browse policy; all the savesets on these tapes will have the same one
    year retention and browse policy.  Some of these savesets will be
    cloned to a backup clone pool, and sent offsite with a one year
    retention.  Other savesets will be cloned to a different backup clone
    pool, and sent offsite with a two year retention.  The backup pools
    will not have their file index entries stored.
    
    Focusing on these two year retention tapes, I'm assuming that, after
    one year, the source tapes (which will be retained onsite in TL8nn
    libraries) will be automatically marked available for writing, and the
    file and saveset information will be automatically purged from the file
    and media indicies.
    
    I'm also assuming that the clone tapes stored offsite will continue to
    live on in the NetWorker media index until the two year retention period
    has passed, at which time the tapes will be marked available.  No file
    or media information will be purged from the media and file indicies,
    because none was created for these clone tapes.
    
    I also believe that, when the saveset is first created, it will have an
    instance value (as reported by mminfo) of 1.  When the saveset is
    cloned, it will have an instance of 2.  After one year has gone by, and
    the source, onsite copy of the data expires, then the instance would go
    down to 1.  True?
    
    Does this all hang together?  Have I made any bad assumptions here?
    
    Thanks!
    tl

	Since browse and retention are based on the client, not the
	pool or group, how are you going to get two different
	retentions for the same clients?

	I know this is not ideal but it can't be changed with
	out a major rewrite and still allow changing the browse and
	retention after the data is saved.

    Roger
    
    Thanks for the response.  I'll admit I'm still struggling with what
    NetWorker provides vs. what I think I want to do.  Given the scenario I
    present in .0, can anyone offer a work-around, or a different way to
    achieve the goal?
    
    Thans!
    tl

savegrp, savefs and save all have a -e expiration option. Unclear, to me, what
that will give you though. Unfortunately -e doesn't exist for nsrclone.

Not knowing how the -e actually works I wild guess/experiment could be:  Set the
default time to the longer period and then use -e on all the normal backups to
shorten the time?...

/Claes