HDP2 Design of Masters/Slaves

to create new topics or reply. | New User Registration

Tagged: , , , ,

This topic contains 3 replies, has 3 voices, and was last updated by  Dave 1 year, 9 months ago.

  • Creator
  • #39222

    Pavel Hladik


    we are using old version of Hadoop 0.21 and want to upgrade to HDP2 production release in a future. I red documentation Hortonworks.ClusterConfigGuide.1.0 but I’m still not sure what will be best layout of Hadoop services for our small Hadoop cluster.

    We have 1x VM master node and 6x HW slave nodes. The master node is running namenode, secondarynamenode and jobtracker now. Each slave node is running datanode and tasktracker.

    My question is how to deploy basic Hadoop services over 6x HW machines acting as slaves and 1x or 2x VM for masters? Could it be like that?

    1. VM: name node
    2. VM: secondarynamenode, jobtracker
    3. HW: datanode, tasktracker
    4. HW: datanode, tasktracker
    5. HW: datanode, tasktracker
    6. HW: datanode, tasktracker
    7. HW: datanode, tasktracker
    8. HW: datanode, tasktracker

    I will really appreciate your help for redesigning concept for new HDP2 platform.



Viewing 3 replies - 1 through 3 (of 3 total)

You must be to reply to this topic. | Create Account

  • Author
  • #39439


    Hi Kait,

    What are you looking to achieve? You can potentially distribute the services how you wish.
    You would be best to read the documentation available and plan your HDP install dependent on the architecture you want to achieve.

    As you say you have an 6 Physical machines and an “unlimited” number of Virtual Machines, I dont fully understand your question.




    Pavel Hladik

    Hi, thanks for reply, but it doesn’t answer my concrete question how to distribute basic hadoop services over six HW machines and unlimited number of virtual machines?


    Koelli Mungee

    Hi Pavel,

    Please take a look at our documentation on this here:

    It shows a layout of the master nodes (HDFS NameNode, Secondary NameNode, YARN ResourceManagers) and slave nodes for a typical hadoop cluster.


Viewing 3 replies - 1 through 3 (of 3 total)
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.