Checkmk

check_mk is a very intuitive, easy to use and easy to setup monitoring system. It scales perfect vertically and also horizontal. it provides already a huge amount of monitoring functionalities for a lot of environments and it is quite simple extensible with own functionalities. Beside state based monitoring it has also a complete solution to integrate event based monitoring. With the Business Intelligence we are able to picture our mission critical processes in monitoring and able to simulate the impact of outages to the business.

The design of the front-end is a bit vintage style and is in hour highly distributed implementation somehow slow.

We are using the Raw Edition, monitoring over 7000 services on nearly 400 hosts. Although we have another product for monitoring switches and routers, we have integrated status updates from that product into the Check_MK view, so we have a single pane of glass. Almost every aspect of the software can be configured through a web page. It's relatively easy to extend and even modify the supplied checks, and if you do, you don't prevent future upgrades, because the upgrade process supports merging of modifications. The agents are read-only, so inherently more secure, and pretty simple to install. The BI feature enables us to provide monitoring of the business services people actually use, as well as the underlying infrastructure. When there are failures, you can drill down to see which component actually failed, and when. The automatic generation of graphs makes it easy to do trend analysis.

I didn't find it easy to understand the way to use tags. The difference between maintenance windows on hosts and those on services is also not obvious, and similarly acknowledgements on hosts and on services. But these are minor issues.

We liked the thousend of customizing possibilities to build the monitoring for our needs.
Another plus is that CMK comes with a lot of checks for different devices included so you can monitor the most things right away.
The new Business Intelligence "module" and the Reporting features is awesome :-)
There is a free & paid version available, so you can start free and upgrade if you need more features.
The new and automatically generated Network Topology is very nice

- Finding the correct ruleset to tune the values for a check is sometimes not that easy :-)
- Handling in general and the UI could be more intuitive/easier
- Troubleshooting is sometimes not easy, and requieres you to be aware of the commandline of cmk and the correct paths to check the logs. (What about a cool Log overview directly in the Web UI? And a log bundle collector to report issues in the forum/support.)