Vanilla 2 plugin not used cinder/ephemeral drives for storing HDFS data

Fix proposed to branch: openstack-ci/fuel-6.1/2014.2
Change author: Sergey Reshetnyak <email address hidden>
Review: https://review.fuel-infra.org/9212

Fix committed in stable/kilo branch

https://review.fuel-infra.org/#/c/9615/

Reviewed: https://review.fuel-infra.org/9212
Submitter: mos-infra-ci <>
Branch: openstack-ci/fuel-6.1/2014.2

Commit: 97b494c3524f0a37ec5a4c1b5df49ff09fef57dd
Author: Sergey Reshetnyak <email address hidden>
Date: Fri Jul 10 08:12:03 2015

Fix problem with using volumes for HDFS data in vanilla plugin

Using 'dfs.namenode.data.dir' param wrong. Should use
'dfs'datanode.data.dir' for storing datanode data

Change-Id: I3e07fcdca34919daad290acb97f9add2a73a6a4e
Closes-bug: #1473320
(cherry picked from commit c9518993fd73c48210f71604c98b65786f5476db)

Verified on 6.1.
Used package:
2014.2.2-1~u14.04+mos17
in mirror:
http://mirror.fuel-infra.org/mos/snapshots/ubuntu-latest/ mos6.1-proposed/main amd64 Packages

Hi, all. I've create 2 different cluster today. The workers from first cluster have 2Gb cinder volume each, and then I create a cluster without cinder volumes. The output from 'hadoop dfs -df -h' was the same.
So, I guess something went wrong and patch isn't landed or don't have effect.

@starodubcevna - please provide more information, at least ISO # and specific steps you performed to see the issue.

@vsedelnik - ISO #265. On Friday I try to use vanill template with 1 master and 3 workers with 2Gb cinder volumes each. @sreshetniak told me that I should try with 10Gb volumes.
Steps are easy to perform (it can be find in upstream bug description):
1. Create cluster with Vanilla 2.x plugin with cinder storage. Last time I use 2 workers with 10 Gb volumes, and 1 with ephemeral drive (lab limitations)
2. When I logged in to the master node:

ubuntu@xxx-vanilla-2-master-001:~$ hadoop fs -df -h
Filesystem Size Used Available Use%
hdfs://xxx-vanilla-2-master-001:9000 112.3 G 217.4 M 101.9 G 0%

3. The same output I have when I run this command on the master node in cluster with 3 workers with ephemeral drives.

That's why I think that the bug isn't fixed.

@starodubcevna please provide /opt/hadoop/etc/hadoop/core-site.xml and /opt/hadoop/etc/hadoop/hdfs-site.xml configs from datanodes.

I checked this bug and came to the conclusion that the error is associated with the wrong build image. On new image this fix works good