PushMon

This nagios monitoring tool is based on a push model.

We use a client daemon written in python that runs nagios checks every given interval.
In the settings file you can define at which interval this client should post it results to the server (statusd) that saves all data in redis.

This model is very scalable and modular. You can run all daemons on separated hosts, or all together on you nagios server, using a small setup. in a little test setup, i was able to serve and monitor 3000 clients.

This whole project is written in python.

There are several plugins included to monitor the results posted to the statusdaemon (statusd).
Included is a little restfull webserver to request the results per host and per nagios check, or all at once.

In the plugins directory a config builder tool is available that generates a nagios config for each host that has posted results.

The results daemon reads the redis database and serves the POST data and host-data collected from a "node-classifier". Both the config-builder and the nagios checks depend on the results server to collect the data they need.


Nagios Plugins included:

check_client_alive.py - Check if the client has been posting results recently enough
check_resultsd_alive.py - Check if the resultsd is reachable
check_statusd_alive.py - Check if the statusd is reachable
check_service_for_client.py - Check if a certain check on a client is OK
check_all_services_for_client.py - Check all combined services on a client

PushMon is open source material so please give me free beer and contact me or send pullrequests on github :)