I'm not sure what you are "exactly" trying to do, but if I
understand you correctly,

You can run bpend_notify on a per *client* basis, bpend_notify
receives exitstatus as the last param.  Based on this you can
conditionally execute the pager.  Look for all error codes under
/usr/openv/netbackup/bin/goodies/bperrcode and see what error
codes are of interest to you.  There are two problems, if the
host is completely unreachable (either network problem or something
like that), or if the host is down, your bpend_notify will never
run.

I personally prefer backup_exit_notify which runs on the master
server and sends notification.  However, this will put additional
load on your master.

Else, run /usr/openv/netbackup/bin/admincmd/bperror -U -problems
-d <current_time> -e <anhour_ago> in a cron on an hourly basis.
Do date arithmetic with PERL and page yourself as you please.

-ravi

On Thu, 26 Oct 2000, Kratunis, Anthony wrote:

>
>I need help with notifications of problems during backups. 
>
>I am told from Veritas support that the only notification functionality that
>exists is when a backup is complete, an email can be sent.  This sends email
>on ALL backups, not just the failed ones, so if I set it up to page me, I
>will get hundreds of pages.  All I want is to be told about problems.
>
>
>Does anyone know of any other notification capabilities that I have not been
>told about
>
>OR
>
>Has anyone already written some UNIX scripts to search the Veritas logs and
>generate notification of some sort based on it's findings?
>
>thanks very much
>anthony
>akratunis AT tiaa-cref DOT org