Networker

Re: [Networker] Identifying backups, policies, and tapes

2004-12-13 17:34:00
Subject: Re: [Networker] Identifying backups, policies, and tapes
From: Dave Mussulman <mussulma AT CS.UIUC DOT EDU>
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Date: Mon, 13 Dec 2004 16:32:36 -0600
Fields, David wrote:

Does anyone know how to write script that can tell me what tapes are
required based on the browse and retention polices for any given server at
any specified point in time?  For example, if you use Full/incremental
backups for all backups, and a browse/retention policy of say, 4 weeks as
our working parameters (in this case, there's no reason for the retention
policy to exceed the retention policy).  If you run a backup on the
following schedule:

  1     2    3    4     5    6   7    8    9   10   11   12
 Mon  Tues  Wed  Thur  Fri  Sat Sun  Mon  Tue  Wed  Thu  Fri
 Full Incr  Incr Incr  Incr Skp Skp  Incr Full Incr Incr Incr

Then, the Full that was taken on the first Monday can't expire until all of
the incrementals that it depends on expire, in this case the incrementals on
steps 2-8.

Basically, what this means is that the tapes used in these backups also
share these dependencies (and usually others).   However, I can't seem to
find a way to see a list of what tapes are needed in a report so that I can
verify when a tape is supposed to be expiring.  I can write a script to
infer the information by figuring out what tapes were used for a the full
and incremental backups, but I would like for the report to be based on what
Networker thinks the are the dependencies involved.  I suppose I could go to
the Networker User program and see what times it would require to perform a
"full restore" on the date of the last incremental of the chain since that
should show me the same tapes, but again, that is not really what I'm
wanting to do.


The best I've found is to run the unix nwadmin GUI and look under the
indexes.  That will show you a tree representation of the savesets for a
file system, and will include the ID.  If you see something that's
breaking your policies (a tape that's not recycling,) it usually helps
to look at the savesets on the tape with an mminfo command, and then
check the indexes for those clients to see why the dependencies don't
make sense.

AFAIK, there's no text way to get the index information that comes up so
quickly in the nwadmin GUI.

Managing clients in and out of groups to maintain media dependencies is
the biggest single timesink I have as a Networker admin.  For example,
if a client always missed your Monday full but was awake for the Tuesday
incremental, every incremental would be based off the original full.
The tapes would never seem to come free.  What I do is track certain
failures, and move clients out of those 'normal' groups into failed ones
until I can make a manual backup and put them back in the rotation.

If Networker had some kind of language that said "If a client hasn't had
a full backup in >7 days, make the next backup a full backup," it would
really save me some time.

Dave

--
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. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=