0

I need to connect to a unix system having Hadoop 2.0.0 database using Eclipse Juno on a Windows system.I tried adding an eclipse plug-in for an older version of Hadoop but when I add Map-Reduce Location, I get the following error :

server ipc version 7 cannot communicate with client version 3 in eclipse

As per some blog results through google, the version mismatch is causing the issue.

Can anyone help?

Please help me find the correct plugin or lead me to where I am going wrong.

Unless I add this plug-in I would not be able to coonect to the database..is there any workaround?

Thanks, Hitz

user2176576
  • 734
  • 3
  • 14
  • 39

1 Answers1

1

Couple of things, Hadoop is not a database, it's an opensource framework for distributed computing. You can directly run MapReduce programs on Hadoop with out an eclipse plugin. Simply package the classes in to a Jar, copy the jar to the unix system and use the below command to run the jar.

hadoop jar <Jar Name> <Name of Main Class> <Input Dir> <Output Dir>

If the version of eclipse you have is not compatible with the version of Hadoop or your eclipse. Check the Link to build your plugin.

Jerry Ragland
  • 611
  • 10
  • 17
  • Hi Jerry Thanks for the clarification and link.I think the second option suits me as I need the "Eclipse Plugin" Also I went through the link : https://docs.google.com/document/d/1yuZ4IjlquPkmC1zXtCeL4GUNKT1uY1xnS_SCBJHps6A/edit?pli=1 and I do not understand why I need to install Hadoop on my Windows machine for the plugin? Thanks again.Hitz – user2176576 Jan 02 '14 at 09:28
  • You don't have to configure it. Download it from Apache and start following the steps in the document. – Jerry Ragland Jan 02 '14 at 12:26