AppDynamics Monitoring Extension for Pivotal Gemfire / Apache Geode


This is an author supported extension. Please reach out to the author of the extension for any issues. The source code of the extension can be found here.

Use Case

This AppDynamics extension allows to retrieve monitoring information from Pivotal Gemfire and Apache Geode.

The extension connects to the JMX manager (provided by the Geode locators) and retrieve the information. The list of available metrics is available here:

List of MBeans




Build the extension

The project is built using Apache Maven. Run the following command once you download the code and are in the main folder.

mvn -Dmaven.test.skip=true install

You find the extension at < target >/

Please note that the tests within this project programmatically create a Gemfire cluster. In order to retrieve the Gemfire dependencies for this test, some setup steps are necessary to authenticate against the commercial repository of Pivotal. Please see the "Development section" if you want to run the build including the integration test.

Install the extension

Copy and extract the in the monitors folder of the machine agent. (This should create a GemFire folder within the monitors folder).

Please ensure that you started the machine agent with high enough metric limit ( Depending on our Gemfire/Geode cluster size and the configuration settings, this extension might create hundreds or thousand of metrics.

Please start your machine agent accordingly:

 nohup java -Dappdynamics.agent.maxMetrics=2000 -jar machineagent.jar &



The configuration file (monitors/Gemfire/config.yml) defines which metrics you want to capture, where your GemFire cluster are running and what credentials you need in order to connect.

Define the server that is used to connect to the cluster. Often you point to your Apache Geode/Gemfire locator.

  - host: "localhost"
    port: 1099
    username: ""
    password: ""
    displayName: "GemFire1"

Configure the metric prefix setting. This allows to define where in the metric tree of AppDynamics.

metricPrefix: "Server|Component: <ComponentID >|Custom Metrics|GemFire"

The metric section provides a pre-defined list of metrics that are gathered. The first sublevel (in this case "system") describes the logical section the metrics should be organized.

The mbean attribute defines the mbean query. Each entry under metrics defines an JMX attribute that should be read.

If you like you can define an optional AppDynamics name, that is the name that the metric should have in AppDynamics. If you do not offer this, the attribute name is used instead.

The optional scale attribute allows to define a multiplication factor that is applied to the read value. This is helpful if the read value is a percentage value.

    mbean: "GemFire:service=System,type=Distributed"
        appDynamicsName: "LocatorCount" # optional: Describes the name within AppDynamics
        scale: 1 # optional: Defines a value that the read value is multiplied with (interesting to stuff like hit ratios)
      MemberCount: # please note that you can configure just the value. Do not forget the colon though.

To remove an attribute or a block, simply remove or comment it.


You can also view a blog post about the extension from the author here



To safeguard the code a local integration is setup with the project. This spawns a local Gemfire cluster and runs the extension to retrieve data.

Thus for the testing phase, Gemfire libraries are needed. Pivotal provides these libraries with their commercial repository at Pivotal created a nice documentation on how these libraries can be retrieved (

For you to run this project you need to follow this guide: Create a Pivotal account, setup the security configuration in maven and you are good to go.



Licensed under Apache License Version 2.0.


Optimizations / ToDos

This section lists possible improvements that could be done:

  • Allow to integrate a backup server (e.g. a second locator) that takes over if the initially given server is not available.