==> On Mon, 12 Sep 2005 11:39:32 -0400, "Mueller, Ken" <KMueller AT MCARTA DOT 
COM> said:


> You could always backup the client (X) under a second nodename (X_ACTIVE):
> Backup node X with your normal management classes/retention.  Backup node
> X_ACTIVE with a management class set to only keep the active copy on your
> special disk storage pool - anything inactive will be expired out.

> Obviously, that means having the client backup twice, so that might not be
> practicle in your environment.  Assuming that is tolerable, does anyone see
> a 'fatal flaw' in that idea?


Dang.  "only" twice the effort is quite a bit better than the
re-scrunge-the-aggregates method.  Heh, Never forget you could just use a
bigger hammer.

You'd want to make sure the _ACTIVE stgpools are on FILE devclasses rather
than DISK, so they can be reclaimed occasionally.  That would be a good
stgpool to have a very very high reclaim threshold; if you're wasting 20% of
your disk on administratively empty space, that would be well worth copying
files around.


- Allen S. Rout