AnsweredAssumed Answered

Remote client configuration for Amazon Elastic MapReduce

Question asked by gabor_makrai on Nov 7, 2012
Latest reply on Dec 31, 2012 by srivas
Hi guys,

I have started a new Job flow (cluster) on Amazon Elastic MR with the following options:
 - Hadoop Version: Hadoop 0.20.205 (MapR M3 Edition v1.2.8)
 - Hive program option
 - "Start an interactive Hive session"

After the cluster is ready for me, I checked MapR Control System for cluster health and I found that everything was all right!

Now, I want to configure my desktop's MapR client to upload data and run job remotely! I've tried the commands from this description: http://www.mapr.com/doc/display/MapR/Setting+Up+the+Client, with the master instance public dns address, but I got the following error message when I've tried

    c:\opt\mapr\hadoop\hadoop-0.20.2>bin\hadoop fs -ls
    2012-11-08 11:04:12,5482 ERROR Client fs/client/fileclient/cc/client.cc:301 Thread: -2 Failed to initialize client for cluster my.cluster.com, error Operation not permitted(1)
    ls: Could not create FileClient

I've configured my client with the following command:

    c:\opt\mapr\hadoop\hadoop-0.20.2>server\configure.bat -c -C ec2-23-23-48-150.compute-1.amazonaws.com:7222

Is it possible to reach the cluster from my desktop somehow?

Thank you very much,
Gabor

Outcomes