[mrtg] Re: Threshold problems with 2.9.10

[Daniel B. Widdis]

Paul,

I've recently tried to implement threshholds in 2.9.10 and ran into some
similar problems as you, but not all of them.

Although the documentation indicates that establishing a ThreshDir will only
cause threshhold scripts to be run once when the threshhold is crossed, this
does not appear to be the case.  They run repeatedly.  I fixed this by
having my script file create its own lockfile and aborting the script if the
lockfile exists.  (The ThreshProgOK script then deletes the lockfile.

It appears (for me) to only run the ThreshProgOK script once.  I have only
encountered the repeated script problem when over threshhold values, however
(and the program appears to assume a threshhold of 0 if none is specified).

I might suggest you establish very high threshholds for those routers you do
not wish to run the script (perhaps a global ThreshMaxI[_]: 100% or more).

Finally, the difference in symptoms may be because I'm on a Linux system and
you're on an NT system.  I think the bottom line is that Threshholds still
aren't quite meeting the design specification and you need to account for
the differences yourself until these little buglets get fixed.

Dan



----- Original Message -----
From: "Paul Lawrie" <pal at townsville.qld.gov.au>
To: "'MRTG'" <mrtg at list.ee.ethz.ch>
Sent: Thursday, March 15, 2001 3:51 PM
Subject: [mrtg] Threshold problems with 2.9.10


>
>
> I am a first time MRTG user who has successfully implemented 2.9.10
> monitoring a large number of NT/unix servers and some routers and
switches.
>
> I have read just about everything that I can find about thresholds but I
> still cannot seem to solve this problem.
>
> It seems as though if I add a ThreshMaxI command to *any* of my targets,
> then each and every other target in the config file will attempt to
execute
> an external script. However, because none of them have ThreshProgI's
> defined, they all fail with a bad command/file type of error. I have also
> specified a ThreshDir and this seems to work fine (MRTG creates files in
> it).
>
> Okay I thought, I'll just add a global ThreshProgI[_] type command which
> will save on some typing anyway. However, every single target then
executes
> this script even if it does not exceed the threshold. I noticed that when
it
> does not exceed the threshold only 2 parameters get passed (target name
and
> the value) but when it is exceeded 3 parameters get passed (target,
> threshold and value). So in the script I basically just ignore any
> invocations that only supply 2 parameters. This is contrary to what the
> manual says, and I am concerned about the performance impact of launching
> this script for every single target every time it is polled. A file for
> every single target is created in the ThreshDir each time the script is
run,
> even if it is not over the threshold. In fact, it gets created even if the
> target has no ThreshMaxI (etc.) commands specified !
>
> Apart from this weirdness, the threshold also triggers every time the
script
> is run, which is also contrary to the 2.9.10 documentation. The flag file
in
> the ThreshDir does get correctly deleted when the threshold drops below
the
> set level and works 100% as expected, but it appears as though it is
simply
> not being checked before the ThreshProgx is being called.
>
> I'm stuck... I'm almost confident that I've read everything there is on
the
> matter but I cannot find a solution. I hope it's just a problem with my
> configuration and somebody can help.


--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi