hadoop installation on ec2

0
votes

I am trying to install a multinode cluster in EC2 by following https://dzone.com/articles/how-set-multi-node-hadoop

Everything seemed to work: I have a namenode and a datanode and I have the following processes running:

namenode: 1389 NameNode,1687 JobTracker, 1590 SecondaryNameNode datanode: 1415 TaskTracker, 1286 DataNode

I could check the namenode status from "ec2-XX-XXX-XXX-XXX.compute-1.amazonaws.com:50070/dfshealth.jsp" and check Jobtracker status : "ec2-XX-XXX-XXX-XXX.compute-1.amazonaws.com:50030/jobtracker.jsp"

The problems came when I tried to check the TaskTracker Status in "ec2-XXX-XX-XXX-XXX-X.compute-1.amazonaws.com:50060/tasktracker.jsp "because I received the message "The page is not available".

I have also tried to run a mapreduce job:

hadoop jar hadoop-examples-1.2.1.jar pi 10 1000000

but I received an error:

"File /user/ubuntu/PiEstimator_TMP_3_141592654/in/part0 could only be replicated to 0 nodes, instead of 1"

Could anyone help me with that?

Thanks

1
hadoopamazon-web-servicesamazon-ec2

1 Answers

0
votes

Have you tried setting the Inbound Traffic in Security Group to accept port 50070, 50030 and 50060?

the hadoop command is expecting a jar task named "hadoop-examples-1.2.1.jar". Check that you have an example jar file for hadoop to run. You may download the jar file at "http://www.java2s.com/Code/Jar/h/Downloadhadoopexamples121jar.htm"