Welcome to the djigger documentation pages.

We hope you’ll find all of the necessary information to make use of our open-source profiler and APM for java. If you have any questions, please use the form form our contact page at http://www.exense.ch/contact . We always value user feedback and we’ll be happy to answer them.

Have fun!

A short introduction to djigger

Djigger is a production-ready performance analysis and monitoring solution for java applications which relies primarily on an advanced round-the-clock sampler and an agent mode for instrumentation. The client contains all the means necessary to connect to a target JVM (standard JMX connection capability, process attach, etc) or to analyze it’s behaviour offline or post-mortem (via djigger’s saved sessions, persistence store or jstack output parser). Store mode is a direct connection to the mongoDB instance used to persist the samples taken by our headless collector. TTL functionality at the threaddump collection level allows the user to monitor his java apps ²⁴⁄₇ without having to worry about running out of space. Scalability and performance is managed by the native capabilities of the underlying mongoDB-backed Persistence Store (via setting secondary indexes on custom attributes and using mongoDB’s scalable collections).

Technical architecture overview (APM mode)

Please note that almost all functionality is also accessible in the “standalone” mode, meaning that you only have to start the client. The client then connects directly to the target JVMs and you don’t have to install a collector nor mongoDB.

Why another profiler and APM?

While we’ve found a couple of open source projects highly interesting, we felt we needed to release djigger because we believe there is simply no other open-source tool that comes close to bringing the same level of functionality for java performance engineering.

Please check out our FAQ section if you’re interested in understanding our engineering philosophy and technical approach.

djigger is the result of years of iterations analyzing the performance of hundreds of applications and pouring that experience right back into the tool in the form of custom-tailored functionality. So whether you want to use it, fork our code or contribute to our project, we’re simply happy to share our experience and give people an opportunity to benefit from it. We believe that’s what open source is about.