Skip to end of metadata
Go to start of metadata

Checking out Ambari source

Follow the instructions under Checkout source code section of "How to contribute" guide.

We'll refer to the top-level "ambari" directory as AMBARI_DIR in this document.

Tools needed to build Ambari

The following tools are needed to build Ambari from source.

Alternatively, you can easily launch a VM that is preconfigured with all the tools that you need.  See the Pre-Configured Development Environment section in the Quick Start Guide.

  • xCode (if using Mac - free download from the apple store)
  • JDK 7 (Ambari 2.0 and below can be compiled with JDK 6, from Ambari 3.0, it can be compiled with at least JDK 8)
  • Apache Maven 3.3.9 or later
    Tip: In order to persist your changes to the JAVA_HOME environment variable and add Maven to your path, create the following files:
    File: ~/.profile



    File: ~/.bashrc


  • Python 2.6
  • Python setuptools - python 2.6: Download or python 2.7: Download and run:


  • rpmbuild (rpm-build package)
  • g++ (gcc-c++ package)

Running Unit Tests

  • mvn clean test
  • Run unit tests in a single module:

  • Run only Java tests:

  • Run only specific Java tests:

  • Run only Python tests:

  • Run only specific Python tests:

  • Run only Checkstyle and RAT checks:

NOTE: Please make sure you have npm in the path before running the unit tests.

Generating Findbugs Report

  • mvn clean install

This will generate xml and html report unders target/findbugs. You can also add flags to skip unit tests to generate report faster.

Building Ambari

Note: if you can an error that too many files are open while building, then run: ulimit -n 10000 (for example)

To build Ambari RPMs, run the following.

Note: Replace ${AMBARI_VERSION} with a 4-digit version you want the artifacts to be (e.g., -DnewVersion=

Note: If running into errors while compiling the ambari-metrics package due to missing the artifacts of jms, jmxri, jmxtools:


[ERROR] Failed to execute goal on project ambari-metrics-kafka-sink: Could not resolve dependencies for project org.apache.ambari:ambari-metrics-kafka-sink:jar:2.0.0-0: The following artifacts could not be resolved: javax.jms:jms:jar:1.1, com.sun.jdmk:jmxtools:jar:1.2.1, com.sun.jmx:jmxri:jar:1.2.1: Could not transfer artifact javax.jms:jms:jar:1.1 from/to ( No connector available to access repository ( of type legacy using the available factories WagonRepositoryConnectorFactory


The work around is to manually install the three missing artifacts:


mvn install:install-file -Dfile=jms-1.1.pom -DgroupId=javax.jms -DartifactId=jms -Dversion=1.1 -Dpackaging=jar
mvn install:install-file -Dfile=jmxtools-1.2.1.pom -DgroupId=com.sun.jdmk -DartifactId=jmxtools -Dversion=1.2.1 -Dpackaging=jar
mvn install:install-file -Dfile=jmxri-1.2.1.pom -DgroupId=com.sun.jmx -DartifactId=jmxri -Dversion=1.2.1 -Dpackaging=jar

If when compiling it seems stuck, and you've already increased Java and Maven heapsize, it could be that Ambari Views has a lot of artifacts, and the rat-check is choking up. In this case, try running

git clean -df (this will remove untracked files and directories)
mvn clean package -DskipTests -Drat.ignoreErrors=true
mvn clean package -DskipTests -Drat.skip

RHEL/CentOS 6:


Ubuntu 12:

Ambari Server will create following packages

  • RPM will be created under AMBARI_DIR/ambari-server/target/rpm/ambari-server/RPMS/noarch.
  • DEB will be created under AMBARI_DIR/ambari-server/target/

Ambari Agent will create following packages

  • RPM will be created under AMBARI_DIR/ambari-agent/target/rpm/ambari-agent/RPMS/x86_64.
  • DEB will be created under AMBARI_DIR/ambari-agent/target

Optional parameters:

  • -X -e: add these options for more verbose output by Maven.  Useful when debugging Maven issues.
  • -DdefaultStackVersion=STACK-VERSION
  • Sets the default stack and version to be used for installation (e.g., -DdefaultStackVersion=HDP-1.3.0)
  • -DenableExperimental=true
  • Enables experimental features to be available via Ambari Web (default is false)
  • All views can be packaged in RPM by adding -Dviews parameter
    • mvn -B clean install package rpm:rpm -Dviews -DskipTests
  • Specific views can be built by adding --projects parameter to the -Dviews
    • mvn -B clean install package rpm:rpm --projects ambari-web,ambari-project,ambari-views,ambari-admin,contrib/views/files,contrib/views/pig,ambari-server,ambari-agent,ambari-client,ambari-shell -Dviews -DskipTests

NOTE: Run everything as root below.

Building Ambari Metrics

If you plan on installing the Ambari Metrics service, you will also need to build the Ambari Metrics project. 


The metrics rpms will be found at: ambari-metrics-assembly/target/. These would be need for installing the Ambari Metrics service.

Running the Ambari Server

First, install the Ambari Server RPM.



On Ubuntu 12:

Initialize Ambari Server:

Start up Ambari Server:

See Ambari Server log:

To access Ambari, go to

from your web browser and log in with username admin and password admin.

Install and Start the Ambari Agent Manually on Each Host in the Cluster

Install the Ambari Agent RPM.




Edit the location of Ambari Server in /etc/ambari-agent/conf/ambari-agent.ini by editing the hostname line.

Start Ambari Agent:

See Ambari Agent log:

Setting up Ambari in Eclipse

After doing the above you should be able to import the project via Eclipse "Import > Maven > Existing Maven Project". Choose the root directory where you cloned the git repository. You should be able to see the following projects on eclipse:

Select the top-level "ambari pom.xml" and click Finish.

Coding Guidelines for Ambari

  • No labels


  1. It appears that ambari-server/../ambari-web/public does not exist, but is required for the build. We've attempted on two branches (trunk, and branch-1.2"

  2. Hi Jay, I'm also facing the same issue.
    Ambari Web has trouble building, even though mvn says it's SUCCESS. I'm looking into it.

  3. I saw this problem on a CentOS 6.3 VM today.

    Ambari Web build was failing with some error message about EmberPrecompiler failure.

    This is what I have done to get it working again:

    1. cd $AMBARI_DIR/ambari-web
    2. rm -rf node_modules
    3. npm install
    4. brunch build

    I don't know if there was a temporary problem with the npm repo or what, but things are working for me now...

  4. 1) I had to do "python > 2.6" on RHEL, like so:

    mvn -X -B -e clean install package rpm:rpm -DskipTests -Pcentos5 -Dpython.ver="python > 2.6" > /tmp/mvnlog 2>&1

    2) And I had to specify the exact RPM path:

    yum install ./ambari-server/target/rpm/ambari-server/RPMS/noarch/ambari-server-1.2.2-SNAPSHOT20130401212524.noarch.rpm

  5. Hi Yusaku, I'm not sure this is the right place to post my question but I cant find a more suitabe spot in this forum. I'm trying to write a simple web manager console for a database and learning ember.js the same time. I think Ambari is a good reference so I looked into the source code, especially the web part. I wonder if you are using any project organization tool with ember.js, as I see require('app') kind of syntax everywhere in the java scripts. What is that or what tool you are using to achieve the more organized multi-file single page ember application, if you don't mind to share. I appreciate your help!

    1. Hi Grace,

      We use an application assembler called Brunch:
      It lets us organize JS files in a hierarchical manner using the require syntax that you saw. Also, it can very quickly compile all the files into consolidated JS files (so that when the browser downloads the web app, it's only downloading just a few files, rather than hundreds of source files). It also serves as a Node.JS based web server that we can use for testing locally with mock data without having the actual backend server running, etc.
      Please note that the version of Ember we are using is v1.0.pre. There have been significant changes since then (especially how the router works in the v1.0 release).
      For any questions related to Ambari code/development, please ask on the ambari-dev mailing list:


      1. Thanks so much for the prompt reply. Will study on brunch. And thanks for pointing me the right mailing list, I'm sure I'll make a good use of out of it soon enough!