logic/feature request for dependencies

[jmoseley at pgtv.com]

Andreas Ericsson <ae at op5.se> wrote:

>If the host and all of its parents are down, notifications will be
suppressed.
>If one of the hosts parents are up, notifications will be sent normally.
>If anything else happens, then that is a bug.

Not so according to the network reachability documentation.  UNREACHABLE
notifications will be sent, if turned on.  It states:

-Nagios will determine that all the hosts "beneath" Router1 are all in an
UNREACHABLE state because Nagios can't reach them. Router1 is DOWN and is
blocking the path to those other hosts. Those hosts -might be running fine,
or they might be offline - Nagios doesn't know because it can't reach them.
Hence Nagios considers them to be UNREACHABLE instead of DOWN.


-UNREACHABLE States and Notifications


-By default, Nagios will notify contacts about both DOWN and UNREACHABLE
host states. As an admin/tech, you might not want to get notifications
about hosts that are UNREACHABLE. You know your network -structure, and if
Nagios notifies you that your router/firewall is down, you know that
everything behind it is unreachable.

If you do have unreachable notifications turned on in the hosts config,
then your only option of controlling/repressing those notifications during
a major outage is via the dependency config.  But the dependency
documentation clearly states that if only one 'parent' of a dependent host
is down, notifications will be suppressed:

-If even just one of the notification dependencies for a host fails, Nagios
will temporarily repress notifications for that (dependent) host.

So, perhaps the solution is to simply disable UNREACHABLE notifications in
the hosts config?  It seems after rereading the documentation that
UNREACHABLE alerts are sent only for those hosts that are unreachable
behind a downed parent device.  I can live with that...

However, I still think rewriting the logic in the dependency section of
Nagios to allow a user to specify whether one or all the 'parents' of a
dependent host must be down before notifications are suppressed would be a
worthwhile request.

BTW, I have used Nagios for several years.  We monitor over 500 hosts and
1,200 services - it's a great product!

James Moseley



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV