[Nagiosplug-devel] some addition to nagios I'm working on
william(at)elan.net
william at elan.net
Thu Jun 15 23:28:36 CEST 2006
I'm working on some additions to nagios code as part of my current
project (actually most of it is already done). If you'd like to see
any/all of this submitted as patches for inclusion in your main
code, let me know:
1. Addition of extinfo structures for hostgroup and servicegroup
(mostly duplicated structures & supporting code for service_extinfo),
the config for them are [not surprisingly] hostgroupextinfo and
servicegroupextinfo. I only need to use hostgroupextinfo myself
so servicegroupextinfo code will not be well tested.
2. Additional target_frame field in all extinfo structures and in
their config file that as you could guess is to be used to
overwrite default "target=_blank" when using extinfo urls.
3. Support for having multiple hostextinfo and serviceextinfo for
the same host/service. This would result is more then one icon
being displayed (makes having both notes_url and action_url
kind-of useless, but oh well...). This is probably more
controverseal extinfo addition then the above as it changes
expected error-condition behavior.
4. New status types "HOST_DISABLED" and "SERVICE_DISABLED" that
with status.cgi shows up as separate field in status totals
and not counted in "All problems". This status is actived
when you manually disable active checks, i.e. expect it to be
down for considerle period or it maybe service that was not
up in the first place (i.e. same services specified for hostgroup
and not every host has all of them). For my scenario I wrote it
so that disabling host would automaticly cause all services for
it to be seen as disabled (even if you did not actually disable
them), many of you probably dont want that. BTW - this addition
I've had in my production nagios for several months now and have
been backporting patch from 2.1 to 2.2 and now 2.3 code. This
likely even more controversial then all extinfo changes...
Note that most of my current code additions are done like:
#ifdef SUPPORT_DISABLE_STATUS
..code
#iendif // SUPPORT_DISABLE_STATUS
For real patches to main code you probably want those ifdef gone
but possibly they could be converted to regular 'if' checking
that particular configuration is activated.
--
William Leibzon
Elan Networks
william at elan.net
More information about the Devel
mailing list