Aprof is a Java memory allocation profiler with very low performance impact on the profiled application that can be used (and is used) on highly-loaded server-side applications in production environment. It acts as an agent which transforms class bytecode by inserting counter increments wherever memory allocation is done and tracks a precise size of allocated objects in bytes. It also keeps limited information about allocation context to aid in finding the memory allocation bottlenecks. 

GitHub Repositoryhttps://github.com/Devexperts/aprof
Public Maven repohttps://bintray.com/devexperts/Maven/aprof (check agent/<version>/agent-<version>-bin.zip for distribution)

Using aprof

Download the latest version of aprof distribution, unzip, copy "aprof.jar" to your application's directory and run the application with an additional JVM argument:

java -javaagent:aprof.jar <your-application-options>

Run it and watch for "aprof.txt" file that will be updated every minute with memory allocation details reports for your application for the last minute (with totals in the separate section at the end) in a human-readable form. There will be somewhat slower start-up time due to transformation and a lot of extra memory allocations by aprof itself, so wait a couple of minutes until your application and aprof warm up (aprof allocates all the memory it needs to operate during start-up and warm-up), start your application's workload, wait some more time until your application warms up under load, and use reports in "aprof.txt" to find memory allocation bottlenecks in your application under load.

To get help on configuration parameters, run 

java -jar aprof.jar

Do not rename agent file "aprof.jar", because it is configured by its name in manifest and will not work when renamed. 

Features overview

Known issues

In the current version (build 31):

Origin and Goals

Java VM has an option -Xaprof printing how many instances of classes were allocated during the lifetime of the application and how much memory they occupied in total and per instance. The option has no performance impact due to the fact that counting takes place during garbage collection. The only drawbacks are occasional overflows of counters and the absence of any information on locations where the object allocations take place.

Aprof aims to overcome drawbacks of Java VM option -Xaprof by:

When to use Aprof

Aprof should be used when the application spends a lot of time in garbage collection. Due to the specifics of Java VM, objects are allocated so fast that CPU profilers are unable to pinpoint the allocation hotspot. The only way is to use memory allocation profilers.

There are many good memory allocation profilers. Unfortunately, in order to find locations where object allocations take place, they do the following:

As a result, memory allocation profilers leave us with both inaccurate profiling results and an impact on the profiled application.

Aprof gets accurate profiling results and finds locations of object allocations at the same time being garbage-free and having a very low performance impact.

How it works

See presentation on Joker Conference 2014: http://www.slideshare.net/elizarov/aprof-jocker-2014