Java is sometimes called the "king of the enterprise." It is so incredibly popular in large systems despite often-cited drawbacks such as memory usage that one might wonder what makes it so attractive. One reason could be that it lowers maintenance costs—at least that is claimed sometimes, and it would make a lot of sense in large, long-living systems. Developing a system is usually cheap compared to maintaining it over a long period of time. Given the widespread usage of Java-based systems, the built-in JMX support is very handy—except maybe the limiting endpoint support. In this chapter, we looked at setting up a separate daemon called the Zabbix Java gateway and performing the initial configuration to make it work with a Zabbix server. We also monitored heap memory usage on the gateway itself, and that should be a good start for JMX monitoring. For easier debugging, we used a simple wrapper around zabbix_get to query JMX through the gateway manually.

Lately, we have been discussing the monitoring of somewhat niche products and protocols. The next chapter will continue that trend—we will discuss the built-in VMware monitoring that enables us to discover and monitor all the virtual machines from a hypervisor or a vCenter.