HDP on Windows – Installation Forum

Running 2 Masters on windows

  • #18241
    rui favas
    Member

    Hello, I’m trying to create an HDFS cluster with 2 masters (2 servers). do you have any guidance on how to run the install?

    e.g.

    server one runs:
    NameNone
    Secondarynamenode
    Datanode

    server two runs:
    NameNone
    Secondarynamenode
    Datanode

    Clients can connect to server one or server two and read/write files which get synced automatically across both servers

to create new topics or reply. | New User Registration

  • Author
    Replies
  • #18254
    tedr
    Member

    Hi Rui,

    The configuration you are trying to set up is not currently supported. You can go ahead and attempt it, but the only configuration pointer I can give is that you will probably need to set up a VIP to switch between them. Also running the namenodes and datanodes on the same server is not reccomended. A similar configuration to this will be supported when hadoop 2.0 is fully adopted, but under hadoop 1.0 it is not. Hadoop 2.0 is currently still in beta.

    Thanks,
    Ted.

    #18255
    rui favas
    Member

    Thanks for the reply Ted.

    Could I run a master datanode on server 1 and a slave datanode on server 2? And if so, would I still be able to write to server 2 if server 1 goes down?

    Do you have any dates for version 2.0 on windows?

    #18282
    tedr
    Moderator

    Hi Rui,

    In hadoop there is no such thing a a "master DataNode" or "slave DataNode." The master for the DataNode is the NameNode. Currently hadoop is made to use only one NameNode and one or more DataNodes. Also there are currently no dates set for release of hadoop 2.0 on windows as it is still in beta on Linux.

    Thanks,
    Ted.

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

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.