12 comments

[ 1.5 ms ] story [ 44.3 ms ] thread
I've had the opportunity to work with Icinga during my internship last year.

Nagios development has stagnated somewhat but Icinga is thriving. It is 100% compatible with all Nagios plugins and most of the configuration tools. Most notably, NConf works very well with Icinga.

They reworked the old Nagios interface and offer entirely new web and mobile interfaces. I must say it's a pleasure to use.

Is it capable of adding new monitoring checks on the fly yet, or storing its configuration in a database? Those are things that I really want.
Not 'on the fly' but very close. Add them, then click the recompile button and it will rewrite the configs and restart the deamon in a matter of seconds.
Nagios development has stagnated somewhat

Understatement of forever.

Nagios is stuck in 1995 and refuses to advance. Cacti is stuck in 1996 and refuses to advance.

We need a unified monitoring/trending/graphing/alerting/prediction platform for everything from servers to application level metrics. No single product or service exists enabling you to actually understand your platform performance.

When I needed an open source monitoring solution some time ago I eventually went with Zenoss. Nagios felt a bit clunky. Zenoss had a (IMO) better structured web interface and also more options for log-filtering/monitoring. Also it's based on Python so easier for me to extend.

The only drawback of Zenoss is that it is a bit of a memory hog, which eventually became a problem as it was running on a cheap VPS...

Icanga does sound promising, as it looks like it redesigned Nagios from a bag of old perl scripts to a better-integrated solution.

This has said "Error establishing a database connection" for a few minutes now. Quite ironic.
I'm guessing this is their first HN slam.

I submitted it because i found very few mentions of it via HN-search.

We are using Icinga to monitor stuff around our office. Its nice and robust, but sometimes you have to work around the architecture to do what you want. Like, maybe you don´t want icinga to poll data, you would rather send data to it.

Anyways, Ive forked the Greplin-Nagios-Utils to address this issue and added some stuff to read from Memcache, now your super importatnt services can supply data to Memcache, Icinga will then do some haxx and poll the data when it pleases, so you will end up with the data in your frontend. Its in an early stage.. But hopefully I will have the time to make it better, faster, stronger. https://github.com/simonjohansson/greplin-nagios-utils/tree/...

You should be able to send data using passive checks. As far as I know, that has always been the model in Nagios.
In that case I should read the documentation a bit better. Thanks.