You can also run python mapreduce jobs using Hadoop Streaming: http://www.michael-noll.com/wiki/Writing_An_Hadoop_MapReduce_Program_In_Pyth on
But again, this operates on hdfs files. We typically wrap a python MR job in two java MR jobs that handle HBase interaction. -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of stack Sent: Wednesday, October 15, 2008 10:12 AM To: hbase-user@hadoop.apache.org Subject: Re: where to find fresh docs 2008/10/14 ROL <[EMAIL PROTECTED]> > > 2) > Does anyone have working mapreduce example for hbase in jython? It depends on what you are trying to do but you should check out http://code.google.com/p/happy/. The documentation in the download is better than what is up on the site. It depends on what you are trying to do but as is, it does not have support for hbase TableInput/OutputFormat nor TableMap or TableReduce but it doesn't look like it'd be hard to add (and I'm sure Colin would take any patch that added such facility). Out-of-the-box, it like it'd be easy writing an uploader in happy if the input was files in hdfs and in your map you did the inserts into hbase. In your subclass of happy.HappyJob, in the __init__, you'd create an HTable instance and then use it at map time doing the hbase updates. This page has jython examples that go against an old version of hbase: http://wiki.apache.org/hadoop/Hbase/Jython. It might help. St.Ack
