How to configure HTTPFS for filebrowser in NamenodeHA cluster

to create new topics or reply. | New User Registration

This topic contains 3 replies, has 2 voices, and was last updated by  GeKo 1 year, 4 months ago.

  • Creator
  • #49607


    I’ve installed the latest HDP2 via Ambari, enabled Namenode-HA and installed Hue as described here.

    I can login to Hue, but the filebrowser isn’t working. In Ambari =>Service HDFS => General WebHDFS is enabled, but none of the cluster nodes is listening on configured port 14000?!?! Therefore it makes sense, that “check for misconfiguration” in Hue shows the error:
    hadoop.hdfs_clusters.default.webhdfs_url || Current value: http://hadoop-pg-7.cluster:14000/webhdfs/v1/ || Failed to access filesystem root

    Why is none of the namenodes offering WebHDFS / listening to 14000 ?!?! I restarted the services several times to ensure the latest config has been applied.

    Additionally I found the information that for environments using Namenode-HA the Hue filebrowser shall use HTTPFS instead of WebHDFS. Is this still correct and how to configure this in Ambari ?!? I didn’t find any hints for that.

    many thanks in advance, Gerd

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

The topic ‘How to configure HTTPFS for filebrowser in NamenodeHA cluster’ is closed to new replies.

  • Author
  • #49894


    that’s perfect, many thanks Dave.



    Hi Gerd,

    In your hue.ini you should have a setting for:


    When using HA you should set this to the service name which you specified in the HA setup (or it is found in the hdfs-site.xml under <name>dfs.nameservices</name>

    This will only point to the service and not worry about going directly to the NameNode






    sorry for bothering 😉
    I thought HTTPFS is also started automatically while restarting all the services via Ambari.
    I manually started this service and configured in hue.ini the current active namenode as defaultFS and now the filebrowser is working.

    There is just one open question: if a namenode takeover occurs the filebrowser would stop working since the configured server isn’t the active one. Do I have to reconfigure hue.ini anytime the active namenode switches over ?!?!

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.