Monitoring plays a vital role in an IT administrator's life. It makes the life of a web/infrastructure engineer predictable. When I started my career in web infrastructure support, I always wondered, how does my boss know that a process is 90 percent utilized for a particular system or how does he know that a particular process will die after about 90 minutes from now, without logging into the application? Later, I found out that they have set up a monitoring system using various third-party tools available in the market for servers and application monitoring.

In this chapter, we will discuss:

Before we learn how to monitor Tomcat 7, let us understand why monitoring is actually required for any system, as we have configured the systems well for the user.

The answer to this question is very tricky. In a real-time environment, the system may break down due to many reasons such as a network glitch, sudden CPU spike, JVM crash, and so on. There are some revenue-generating applications, for example, if bank sites go down, then there will be a huge revenue loss, also administrators will not know unless users start complaining about the issues. This will also have a bad impact on the business. If monitoring systems were set up on the server, the web administrator would get a notification stating that the following systems are going down, and he/she would take the necessary actions to fix the problem. Hence, it minimizes the impact on the application downtime.

Different ways of monitoring

In today's world, with increasing infrastructure, it becomes very difficult for administrators to manage servers. In order to identify the issue beforehand, and to minimize the downtime, monitors are configured on the system. We can configure multi-level monitoring on the systems, based on the infrastructure requirements for example, the OS, Web, Application, and Database level servers and individual application level. There are different ways of configuring multi-level monitoring. The following figure shows different ways to configure monitoring for any infrastructure:

Monitoring can be mainly done in three ways on a system, which are as follows:

• Third-party tools
  • Monitoring setups are configured using third-party tools present in the market, such as Wily, SiteScope, Nagios, and so on.
  • These kind of monitoring tools are used in an enterprise infrastructure setup, where there are more than 100 servers with different infrastructure components (domains) such as web, application, database, filesystem servers, and so on.
• Scripts
  • Scripts are used in monitoring, where a specific use case needs to be monitored, such as the results of how many users are logged in for a particular interval of time or application-specific user roles.
  • Used everywhere in small and big IT organizations.
• Manual
  • This process is used when any application's performance is slow for a particular module.
  • Mostly used at the time of troubleshooting and where the number of systems are less than three.