PDA

View Full Version : Reading from HBase in a mapper transformation



markob
07-03-2013, 09:52 AM
Hello,


I am trying to read from HBase in Kettle 4.4.0 (29 October 2012) in my mapper transformation using MapReduce Input and HBase Row Decoder by using this example job/workflows:
http://jira.pentaho.com/browse/PDI-7182?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel
I get the following error when executing the job:


No Table was provided


Seems that the mapper / job is unaware of the table name, which is set both in HBase Row Decoder and in Job config dialog as 'hbase.mapred.inputtable' user defined variable.
Either the ISSUE (above link) is not solved or my Kettle version is wrong.
The host hadoop is CDH4.2.1 with HBase 94 and I have the following jar's:
./plugins/pentaho-big-data-plugin/hadoop-configurations/cdh4/hbase-0.92.1-cdh4.0.0.jar
./plugins/pentaho-big-data-plugin/hadoop-configurations/cdh4/pentaho-hadoop-shims-cdh4-1.3.0.jar
Also, reading from HBase from HBase Input step in a non-mapper workflow works.


Thanks