Saving 20 or 30 characters won't do much to solve the detail output
limitations
of monitoring applications such as nagios. I have some checks that generate
100+ lines of output. What I'm doing is to store the detail output in an
array
and save it to a file, then include a link to the file in the normal
plugin multi-line output.
Lately, I've gotten a little fancier. I have two files, one for 'ok'
output, and another
for 'problem' output, and have two links in the plugin detail
output...one for
the current output, and another for the last 'problem' or 'ok' output
depending
on the current state of the check. That way, if the last problem has already
cleared before I get around to looking at it, I can see the last non-ok
details
and if there is currently a problem, I can see the last 'ok' check
output before
the current problem occurred.
I know this is not very portable, but it represents another way of
thinking
about managing plugin output.
/usr/local/nagios/ssi/results/services/ok/<servicename>/<hostname>.html
/usr/local/nagios/ssi/results/services/problem/<servicename>/<hostname>.html
This helps users to diagnose problems, particularly for plugins that check
several things at once, like all the volumes on a NetApp cluster, or all
the VIPs on an F5 BigIP.
On 10/14/2015 6:01 AM, Holger Weiß wrote:
I definitly like the SERVICE STATUS - OUTPUT|PERF format most.
It seems I'm the only one who doesn't like it
<https://www.monitoring-plugins.org/archive/devel/2007-July/006121.html>
:-) I should just make this configurable
<https://github.com/monitoring-plugins/monitoring-plugins/commit/006597a12d28d9cd752b3eeb56dff3bc97049f4c>
with a command-line flag one day, without changing the current format
by default.
From time to time i have to run the plugins manually from the
command line and it definitly helps.
I see your point for the |STATUS| part, but the |SERVICE| part helps
as well?
—
Reply to this email on GitHub
<https://github.com/monitoring-plugins/monitoring-plugins/issues/1380#issuecomment-147997825>.
