First, we learned to monitor web pages based on various parameters, including response time, transfer speed, HTTP return code, and text, contained in the page itself. We also found out about how to set up multiple scenarios and steps in them as well as setting up variables to be used in all steps. As a more advanced example, we logged in to the Zabbix frontend and logged out of it. For that to work, we extracted the session ID and reused it in subsequent steps. With this knowledge, it should be possible to monitor most of the functionality web pages have.

For production systems, there usually will be way more applications, scenarios, and steps. Web monitoring can be used for many different purposes, the most popular being site availability and performance, but there are many different cases one could monitor, including things such as watching the Slashdot front page for a company name and replacing the usual first web page with a more simple one to withstand the coming load—slashdotting—easier.

As a simpler alternative, we also explored web page items on the agent side. They have three features:

Web scenarios are only available on the server side, while the simpler items are only available on the agent side.

Having mostly concentrated on Linux system monitoring so far, we'll depart from that in the next chapter and look at Chapter 14, Monitoring Windows. We'll look at the native agent for Windows, performance counter and Windows Management Instrumentation (WMI) monitoring, and service discovery and Windows Event Log monitoring.