Trouble to create a new table using HCat under HUE

to create new topics or reply. | New User Registration

This topic contains 3 replies, has 4 voices, and was last updated by  Alex Miller 7 months, 1 week ago.

  • Creator
  • #46167

    Ricardo Gaspar


    I’ve just installed my Hadoop 2.0 using the Automatic way.
    I’ve installed a cluster of three nodes, one master, and two slaves. So in order to proof if my solution works, I installed HUE as the documentation says, and follow the tutorials that there are in Hortonworks pages.

    I uploaded the file, and when I try to create a new table using the HCat, it ends with and error saying:

    IOException: Failed to replace a bad datanode on the existing pipeline due to no more good datanodes being available to try. (Nodes: current=[,], original=[,]). The current failed datanode replacement policy is DEFAULT, and a client may configure this via ‘dfs.client.block.write.replace-datanode-on-failure.policy’ in its configuration.

    Someone could help me.

    Thanks a lot

    best regards.


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

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

  • Author
  • #59019

    Alex Miller

    @teknetik I can verify that your solution resolved this same issue for me with a single-node cluster. The Block Replication setting is under Advanced in the HDFS config. Thanks for your help!



    @Ricardo, sorry this will be a bit late to help you but for future reference and to help others:

    After building from Ambari I had the same issue, a little research brought me to this thread among others.

    It seems that Haddop will default to trying to replicate to 3 datanodes. While Ricardo states he has 3 node “cluster” I am guessing from his logs only 2 are datanodes.

    For a manual in stall edit the hdfs-site.xml or via Ambari and look for the Block replication vales and change it to the number of data nodes, in Ricardos case (and mine) 2. Restart your services, Ambari offers to do this for you (I love Ambari!)

    I could then continue with the tutorial.

    {NOTE: Accuracy of this info cannot be guaranteed as I am just starting out with Hadoop myself}



    Hi Ricardo,

    Can you check the datanode logs for errors when you get this error in HUE. Can you create a new table in HCat outside of HUE?



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.