[Nagios-users] Fwd: NSCA MAX_PLUGINOUTPUT_LENGTH

[Ben O'Hara]

On 11/1/05, Andreas Ericsson <ae at op5.se> wrote:
> Please don't crosspost to everywhere. It's about as polite as knocking
> on a door with the business end of a shotgun.
>

Appologies, thought it was revelent to both devs and users!

> I've moved this to nagios-devel only for now, since that's where it belongs.
>
> Ben O'Hara wrote:
> > Ok,
> >    So looking into this it appears that the limit is actually in
> > nagios deinfined in includes/common.h as 348.
> >
> > Obivously with more data being presented in the output, including
> > perdata this is rather small.  Im having issues with getting perfdata
> > for a "DISK" check that checks all filesystems so the output ends up
> > being rather large.
> >
> > Anybody see any reason not to raise this to a higher value? 2048? and
> > also change it for NRPE?
> >
>
> For nrpe and nsca, no. For Nagios it could mean service-checks are being
> garbled since the maximum POSIX-guaranteed atomic write-size is 512
> bytes. That must also include hostname and service-description, so 384
> is actually a calculated value for MAX_PLUGIN_OUTPUT_LENGTH.
>

OK, didnt know this, although it does appear that hostname and service
description are seperate variables with their own defined length.

Im not a coder though so not 100% sure whats going on.

 > > Does this effect other people? Should it be made as a generic
change to nagios?
> >
>
> It shouldn't be made a generic change to Nagios. The change that
> *should* be made is that Nagios should fork off commands and multiplex
> the output so as to get rid of the bottleneck that the FIFO introduces.
> I've done some fairly serious work on this but it sort of dropped
> between chairs as I got busy with other things (which is really a shame
> since it could easily provide a boost in nagios' performance by a factor
> of 40).
>
> I'll get working on it again some rainy day or so.
>

Great, seems a shame that the output is limited as obviously this
causes problems with perfdata etc, especially when its related to disk
usage which could produce a very large output.

Regards

Ben

> --
> Andreas Ericsson                   andreas.ericsson at op5.se
> OP5 AB                             www.op5.se
> Tel: +46 8-230225                  Fax: +46 8-230231
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by the JBoss Inc.
> Get Certified Today * Register for a JBoss Training Course
> Free Certification Exam for All Training Attendees Through End of 2005
> Visit http://www.jboss.com/services/certification for more information
> _______________________________________________
> Nagios-devel mailing list
> Nagios-devel at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/nagios-devel
>


--
"There are 10 types of IT people. Those who understand binary and
those who don't."


-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.
Get Certified Today * Register for a JBoss Training Course
Free Certification Exam for All Training Attendees Through End of 2005
Visit http://www.jboss.com/services/certification for more information