perfino is a monitoring tool for the JVM. It is intended for in-production use and adds extremely low overhead to monitored applications. Its mode of operation is characterized as APM, short for "application performance management". Rather than collecting performance data at a low level and with a broad scope, perfino presents selected operations at a high semantic level, called "business transactions". In addition, scalar data is monitored from a variety of sources. Based on that data, threshold violations can result in alerts that help you safeguard the quality of service of your applications.
perfino is intended to run with your application at all times. This enables it to focus on historic data, showing you how performance characteristics evolve over long periods of time. Data is automatically made less granular as time goes by, so you can look back years into the past with only a slow rate of storage space consumption.
perfino is designed to monitor multiple VMs and trace the interactions between them. Whether you have a number of fixed VM installations or a cloud deployment with hundreds of VMs, perfino can monitor and organize them at the same time.
The perfino UI is a web interface that can be used by multiple users to analyze the collected data at the same time. A system of access levels allows you to partition a single server for multiple groups.
This documentation is intended to be read in sequence, with later help topics building on the content of previous ones.
First, a broad overview over the architecture will help you to understand the components of perfino.
Subsequent chapters build your expertise with respect to different functionality in perfino. The "Configuration" and "Advanced topics" sections are optional readings that should be consulted if you need certain features.
If you are using the keyboard for navigation, press
F4 to switch between the index and the
content area. The keys
k move between help topics and the keys
d move up and down when there is a hierarchy of chapters.