> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> Steve Burton wrote:
>> I have nagios 3.0.4 running on FreeBSD and I'm using nagiosgraph to
>> generate rrd's on the data. I'm visualizing the data with both
>> nagiosgraph and drraw.
>> My problem is that the graphs have gaps so I guessing that the rrd's
>> are not being updated. I'm concentrating on just one service for
>> debugging and I've run the plugin (which I wrote) from the
>> command-line many times for a local server and I haven't seen it fail
>> to connect or report or to time out. It is as though nagios is not
>> checking the service within the heartbeat of the rrd's in question. I
>> have set the service check interval to 5 minutes, which I believe to
>> be the default nagiosgraph step time and the heartbeat is set to the
>> nagiosgraph default of 600 seconds.
>> Can anyone advise me if if this is likely to be the problem and if not
>> how I can diagnose what is actually going on?
> I have put a diagnose on the mailinglist for something that is either
> the same problem or something similar.
> The problem is that nagios can send several commands into the pipe at
> once and the other end will only pick up one of them. So the RRD info is
> not written. The work around is not to allow Nagios to run things in
> parallel. But it will severly limit the amount of hosts and services one
> can check with Nagios.
> This should have been fixed by using the file interface instead of the
> named pipe. The file interface was introduced about a year ago or even
> longer back.
> Please check the mailinglist archives for more details.
I have the following in my nagios.cfg
# grep service_perfdata ../nagios.cfg|grep -v #
Also, the perfdata.log file size changes with time, all of which seems
to suggest to me that I'm using the file interface. Is this correct?