1

I have installed Hadoop 2.5.2 single node on Ubuntu 14.04 using this guide http://www.vasanthivuppuluri.com/hadoop/installing-hadoop-2-5-1-on-64-bit-ubuntu-14-01/ when i enterstart-all.sh then jps output doesn't show datanode process.

hduser@jerrygoyal-Lenovo-B490:~$ jps
18043 NodeManager
18327 Jps
17913 ResourceManager
17763 SecondaryNameNode
17018 NameNode

i have also tried methods given on Datanode process not running in Hadoop but no progress.

Community
  • 1
  • 1
GorvGoyl
  • 42,508
  • 29
  • 229
  • 225

2 Answers2

2

Datanode will not run because of several reasons. Some possible solutions that may help you are.

1) Check the datanode logs. You will get the exact reason for failure

2) Clear the datanode storage directory in your set up and restart all the services. If the datanode directory is already having an entry with a different namespace id, it will not start.

3) Check the permissions of datanode storage directory. If the hadoop user is not having proper read/write permissions, it will not run. If this directory is not created, create it and give proper permissions.

You can locate the datanode directory from the dfs.datanode.data.dir property in your hdfs-site.xml.

Amal G Jose
  • 2,486
  • 1
  • 20
  • 35
0

The problem would be the invalid clusterID. for that remove all the folders available in dfs.datanode.data.dir in your hdfs-site.xml

Then start your datanode.

Rengasamy
  • 1,023
  • 1
  • 7
  • 21