Nagios Support Forum

Thanks for the patches Stephan - I've finally gotten around to
committing them to CVS. The one change I did make had to do with the
additional NEB event type. Instead of creating a new event type, I
simply used the NEBTYPE_TIMEDEVENT_SLEEP event type, which had not
yet been implemented.

On 19 Sep 2005 at 11:04, "P=F6tschner, Stephan" wrote:

>
> Hi Ethan, hi community.
>
> I am currently working on a Nagios module and I=B4d like to add some
> comments and code.
> But first of all I=B4d like to thank you for this great software, you
> chose to distribute as open source.
>
> *) The project - I=B4ve already mentioned - is based on C++ code and it
> would be really nice, if all Nagios header file would be guarded by a
> conditional `extern "C"=B4 statement. This way one can use every Nagios
> function in its module code, just add the appropriate headerfile and
> compile the code even with c++ compilers without any problems.
> I=B4ve already added this directive to your header files and created a
> diff-file out of the code (`patch_include=B4)
>
> *) The second thing I would consider as rather helpful, would be an
> additional event, which fires if both queues - high and low priority -
> are empty and Nagios plans to change into idle mode (sleeping for
> `sleep_time=B4 seconds). This feature would be really nice, since one
> could add additional customized tasks or analysis using the NEB
> interface.
> I=B4ve therefore added an additional eventtype:
> `NEBTYPE_PROGRAMSTATUS_EMPTYQUEUES=B4
> And a few lines of code in `events.c=B4. Like the original code, first
> the high priority event queue will be checked, if it=B4s empty the low
> priority queue will be checked, if it=B4s empty, the
> `NEBTYPE_PROGRAMSTATUS_EMPTYQUEUES=B4 event will be fired, so modules
> can do some additional work. Then the check for external commands
> will be executed and at last - if in the meantime no other event
> reached its deadline - Nagios may sleep for some time.
> I=B4ve got one question concerning this context: in "events.c" line
> 906: when will this condition fire (or will this condition ever
> fire), since the service reaper event or the log rotation event
> should be in the queue, even if there are no active services.
>
> *) I=B4d also appreciate, if one could add services that are disabled
> by default and may be executed by NEB modules if needed. I thought of
> using the `normal_check_interval=B4-parameter for this function, since
> your code already seems to catch services with interval `0=B4 after
> they got executed and disables them again.
> This would be very helpful for my project, because my module wants to
> add from time to time an additional service which is supposed to be
> executed exactly one time by nagios and then should be suspended
> again. Therefore I just had to change the sanity check in `objects.c=B4
> to accept intervals with a value set to 0 and did some minor
> additional changes to ensure that this service will be disabled
> again.
>
> I hope you consider my comments as helpful and accept to add them to
> the community release, since I consider your module architecture as a
> very powerful and dynamic method to add new features and would really
> like to use it in my project the way I described.
>
> If you think I could add these features in a better way, please don=B4t
> hesitate to comment on them - I will be glad on any comments in
> general.
> The tgz-archive contains 3 files. One for every folder I added code
> to (patch_base -> ./base, patch_common -> ./common, patch_include ->
> ./include).
>
> Best regards,
> Stephan P=F6tschner
> Research & Development
> Fabalabs Software GmbH
> Honauerstra=DFe 4
> A-4020 Linz
> Tel.: [+43] (70) 60 61 62
> Fax: [+43] (70) 60 61 62-609
> E-Mail: [email protected]
> Web: http://www.fabalabs.org
>
>
>



Ethan Galstad,
Nagios Developer
---
Email: [email protected]
Website: http://www.nagios.org






This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]