Some of the tools in the tuning area have started out with support for one thing or, perhaps, for a few things such as memory analysis or load testing. Over time, however, quite a few of them have evolved in their own right or with the aid of plugins to often have multiple functions. They also often cover several areas in the tuning field.

So, what kind of support do we expect from a set of tuning tools? When looking at the enterprise stack, it becomes clear that we need different types of tools depending on the different layers of the stack:

In general, profiling a software system involves analyzing the behavior and performance usage of resources that are involved, as the system executes. This analysis is most often performed to find points of optimization and for the removal of bottlenecks in the system. Profiling is also commonly, but not necessarily, performed under various levels of load on the system.

Profiling is a very powerful function for investigating and helping resolve suspected bottlenecks. Profiling can also be very complex, and the wrong interpretations may lead to disastrous results. It should, therefore, only be performed in stable environments, preferably, by staff that has shown an aptitude for it.

In the Java universe, profiling technically involves instrumenting the Java bytecode in order to make accurate calculations of execution down to the method level. Instrumenting the Java bytecode is a time-consuming operation that is performed for all code in the JVM when profiling starts. The amount of instrumented bytecode can, however, be limited by defining what should be included or excluded, thereby speeding up the instrumentation time in profiling. Depending on tool support, theses definitions can be done on package, class, or method level. As the profiling terminates, the bytecode is normally restored.

Sampling is a kind of lightweight profiling. Instead of physically instrumenting the Java bytecode as done in profiling, sampling relies on analyzing the stack traces and thread dumps that are taken at regular time intervals. These intervals can be defined, for example to 10 per second.

So, what approach should you use? As always, the answer is that it depends on...

A very useful and free JVM profiler is VisualVM. At the time of writing this book, this tool normally comes bundled when you download and install the Java Development Kit (JDK) from the Oracle Java site (http://www.oracle.com/technetwork/java/javase). In this distribution, the tool is located in $JAVA_HOME/bin and named jvisualvm.sh (for UNIX-based environments) or jvisualvm.bat (for Windows).

As an alternative, you can download it from its own site (http://visualvm.java.net). This is where you'll find the unbranded bleeding-edge version.

The core of VisualVM is shared by the NetBeans IDE (https://netbeans.org). So, if you use the NetBeans profiler, you'll be using a lot of the functionality that is the same as those found in VisualVM, but with differentiating workflows. There's also a launcher plugin of VisualVM for the Eclipse IDE (http://eclipse.org) available at https://visualvm.java.net/eclipse-launcher.html.

All examples in this book are based on the current version, VisualVM...

Keeping track of how your system reacts to different types of influences is imperative when you're testing it, tuning it, and over time in production. The measured values are extremely useful for both technical and business trend analysis as well as for comparing the performance between releases. When done correctly, these trends and comparisons can save your organization a lot of headache and enable it to quickly predict and adapt to upcoming problems.

There are explicit tools that monitor specific features and services, and there are others that give a more collective view of system health and performance. As we have seen, tools such as VisualVM can provide us with a multitude of information from various layers in the enterprise stack. It will give you the basic information about the hardware in terms of CPU usage, and as expected, it will provide a lot of information about the JVM as well as the middleware and the execution of inherent applications within the JVM.

A system that...