HDFS Forum

Add master services like HBase Master or Zookeeper after successful deployment

  • #14938
    Francois BORIE


    Firstly, many thanks for providing such a great hadoop platform as HDP 1.2 ! We appreciate that.

    My problem is the following :

    I’ve successfully deployed HDP 1.2 using Ambari.

    But my cluster has grown since this time and I would like to add more master services like Hbase masters, Zookeeper
    masters … to my cluster.

    Is there any way to do that through Ambari, or directly with shell commands. Maybe I miss something, but it seems to me
    I didn’t see any documentations about that.

    If it’s not a feature available for the moment, may you provide a release date for this feature ?

    Many thanks,



to create new topics or reply. | New User Registration

  • Author
  • #14943
    Sasha J

    Ambari is not support services addition in current release…
    It should be implemented in future releases, but not now…
    However, you can add binaries and start services manually, it will not be visible by current Ambari implementation, but you will be able to start/stop it manually and use.

    Thank you!


    Hi Francois,

    Thanks for your kind remarks about HDP 1.2 and Ambari.

    I am looking into this issue, but it looks like you are correct that there is currently not a method to add new master services to an already installed cluster, only slave or client services. If I find differently in my testing I’ll post again.

    Thanks again.

    Francois BORIE

    Thanks for your answer !

    For your information, I add ” manually ” a zookeeper master service on a node using your existing Puppet manifests. This works good, but the problem is that dependencies for other services like Hbase (in the case of Zookeeper) are not taken into account in a sustainable manner.
    Even if I modify Hase configuration file to add my new zookeeper host, as soon as I restart the service with Ambari, my modifications are lost because I think valid configuration is stored in Pgsql database and Puppet manifests are generated on the fly and based on these data. This way of working sounds good for me, because you ensure you always get valid configurations, but it prevents all kind of local configuration modifications.

    Thanks again and have a good day !


    Larry Liu

    Hi, François

    The feature is in roadmap. Please keep checking.


The topic ‘Add master services like HBase Master or Zookeeper after successful deployment’ is closed to new replies.

Support from the Experts

A HDP Support Subscription connects you experts with deep experience running Apache Hadoop in production, at-scale on the most demanding workloads.

Enterprise Support »

Become HDP Certified

Real world training designed by the core architects of Hadoop. Scenario-based training courses are available in-classroom or online from anywhere in the world

Training »

Hortonworks Data Platform
The Hortonworks Data Platform is a 100% open source distribution of Apache Hadoop that is truly enterprise grade having been built, tested and hardened with enterprise rigor.
Get started with Sandbox
Hortonworks Sandbox is a self-contained virtual machine with Apache Hadoop pre-configured alongside a set of hands-on, step-by-step Hadoop tutorials.
Modern Data Architecture
Tackle the challenges of big data. Hadoop integrates with existing EDW, RDBMS and MPP systems to deliver lower cost, higher capacity infrastructure.