Dave: unfortunately my environment isn't that simple. I have plenty of
clients
that don't use "All" savesets. I also have funny scheduling
requirements, some groups
that run daily, some weekly. also different backup levels on different
days.
a mess, in other words.
in the meantime I think I'm making a bit of progress on this. it seems
that
if I run "/usr/bin/savegrp -p GROUP-NAME" I get a projection of what
will/should
happen the next time the group runs. I think I can save that output to
a
file one day, run mminfo the next day and save that output, I should be
able
to match merge the two. if my Perl is up to it, that is to say.
as far as "silent failures" go I'm probably the poster child. I get
more
of them that you can shake a stick at, and believe me that's exactly
what
I'd like to do. a short list of some of them would include:

  - saveset was never started
  - aborted due to inactivity
  - remote host refused an attempted connect operation
  - (emergency) Bad or missing record
  - RPC error, Timed out
  - a remote host did not respond within the timeout period
  - RPC error, Can't encode arguments
  - RPC error, remote system error

I know you didn't respond to this just to hear me complain but some
days I
just can't help myself.
I don't make much use of the saveset completion reports because I've
found them
too voluminous to deal with. the reason I'm trying to write the script
is to get
something shorter and easier to interpret.

Jim Lane
Sr. Technical Consultant
Network Services
Toronto Hydro
office: (416)-542-2820
cell: (416)-896-8576

>>> Dave Mussulman <mussulma AT cs.uiuc DOT edu> 31-Dec-02 11:29:15 AM >>>
On Tue, Dec 31, 2002 at 10:31:40AM -0500, Jim Lane wrote:
> I'm trying to write a script that could be run daily to report on
what
> Networker has done during the past 24 hours and compare that to what
> should have been done, IE. what filesystems would have been backed up
if
> all the scheduled groups had run 100%. the first part is fairly
easy.
> I've got a little Perl script that parses the output of mminfo,
sorting
> it in order and improving the format a little. I'm not sure how to
> proceed with the second part. how can a script know everything that
was
> scheduled to have been done by Networker? the UNIX GUI has a
function
> called "preview" under group control that's probably something like
what
> I want. it lists out what a group will do the next time it runs. I'm
not
> sure if there's a way to get this information from the command line.
has
> anybody done anything like this? is there another, simpler way to get
at
> the information I'm looking for here? what I want is a script that
can
> be given to Operations where they can run it and flag cases where
things
> don't get backed up by Networker when they should have been.

Jim,

If you use 'All' savesets for your clients, I would imagine (hope) the
easiest way to find out what didn't back up would be to review the
Savegroup completion notification.  Usually the failures are listed at
the top.  We have the notification sent via email, and then parse it
through a perl script to generate a report (clients with large
backups,
open files, client failures, etc.) and it emails that too.

Every once in a while we'll get a 'silent failure' -- where Networker
will only backup the index for the client, and nothing else, and not
flag it as a failed.  That's something our parser grabs as well.
(This is 6.0.2 ... Anyone know anything about these silent fails, or
if
they've been fixed/addressed?)

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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=