What would be the right management tool for my network? I take a look on Cacti, Nagios, OpenNMS, Munin & Monit ... but I'm not sure which will be the best direction.
I prefer something that will out of the box cover much as possible.
I need to monitor:
Services like: nginx, unicorn, Tomcat, Jetty, MySQL, Elasticsearch, Hadoop, PHP, Comet
Nice to have: Graphs for Bandwidth usage, nginx req/sec
Any proposals?
I like OpenNMS, Observium and Monit used together for their respective strengths.
OpenNMS has excellent polling, up/down and alerting capabilities. The alerts, scheduling and provisioning are extremely powerful. It's useful for receiving traps.
Observium has the prettiest graphs, but doesn't provide the alerting features. It's great to be able to step through Observium's detailed inventory display as well. It's geared more towards networking equipment.
Monit is good for application and service-level monitoring. I use it to make sure that critical services are running on Linux servers. Ensure ntp is on. Ensure crond is running. Make sure sshd is on. If sshd dies (crash, OOM, etc.), restart it.
With the equipment and applications you've listed, OpenNMS would cover it all. Unfortunately, the graphs are of poor quality. Monit would only benefit the Linux servers. Observium would detect and graph everything, but wouldn't provide anything more than up/down notifications to the SNMP syscontact.
