Okay, after reading everyone's input on my original posting, I think I
understand what's happening here -- most of the pieces/parts anyway --
with the client id being stored in two places (database and res file)
and the importance of those values matching, but I wanted to run the
following scenario past the news listing to see what people think. Just
want to get some feed back on these points here since may be way off
base!

SCENARIO: Let's suppose you have a client named fred, and fred gets
backed up for like a year or so, and then one day fred is yanked off the
network and surplused. Now, maybe you keep the index (or maybe not) and
then later at some point a new machine gets installed on the network
with the same host name fred, and you add the new fred to the client
listing. Now, this new fred is gonna have a different client id, by
default, unless you specifically go in and change it back. This scenario
is a little different than the one wherein you just deleted the client
and now you wanna recover older data, and it's not really the same as
the one where a client is renamed, so a few questions come to mind:

Question 1: Should you change the id back to match the original as soon
as son of fred is first created?

I think it makes sense to just leave it as the new id because if you
ever wanted to recover savesets based on the old id, you could just
change it back, recover what you wanted, and then change the id back to
the new one once you're done as long as you did so before the next
backups run for fred. The only danger I can see with this is that you
might forget to change it back in which case the next backups would use
the old client id and then you'd have a little bit of an aberration.

Observation 1: Regardless of whether you should or shouldn't reuse the
same id, it seems logical that as long as the current client id matches
the one for the saveset(s) that you specifically want to recover, then
you should be fine to recover them whereupon the id SHOULD then be
changed back, if necessary, to match the client id for the most recent
savesets.

Observation 2: If you're adding the new fred relatively soon after the
old one goes away then seems reasonable that you would just keep the
original index so if you did need to recover anything recent that was
backed up on the old fred, you could use nwrecover. For example, maybe
you move all the data from the old fred to the new fred, and everything
on the new host is set up the same way with same OS, etc. BUT, if you're
adding the new client long after, or if there really is no association
between the new fred and the old one whatsoever, except for the fact
that they just happen to have the same host names, then seems reasonable
to just create a whole new index and not use the old one at all.

George

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=