Setting up and configuring the Pentaho node dist, Kettle (PDI) and Reporting
Preconfigured Packages
These instructions are specific to the MapR distribution of Hadoop, if you are not using MapR, go to the Configure Pentaho for Cloudera and Other Hadoop Versions page.
Client Configuration
MapR Client
- Follow installation instructions provided by MapR for your architecture: Setting up the Client - MapR
PDI Client
- Download and extract Kettle CE from the Downloads page.
- Configure PDI Client for MapR
- Overview:
- The MapR native libraries for your architecture must be added to the
java.library.path
- MapR Hadoop Configuration directory needs to be on the classpath
- MapR Hadoop Core library must be on the classpath
- The MapR native libraries for your architecture must be added to the
- All architectures
- Update the $PDI_HOME/launcher/launcher.properties with the attached launcher.properties
- Delete $PDI_HOME/libext/pentaho/hadoop-0.20.2-core.jar
- Copy $MAPR_HOME/hadoop/hadoop-0.20.2/lib/hadoop-0.20.2-dev-core.jar into $PDI_HOME/libext/pentaho
- Copy $MAPR_HOME/hadoop/hadoop-0.20.2/lib/maprfs-0.1.jar into $PDI_HOME/libext/pentaho
- Linux x64
- Update the $PDI_HOME/spoon.sh with the attached spoon.sh
- Mac OS X 64-bit
- Update the Data Integration 64-bit.app/Content/Info.plist with the attached Info.plist
- Overview:
Report Designer
- Download and extract PRD from the Downloads page.
- Configure PRD for MapR
- Delete $PRD_HOME/lib/jdbc/hadoop-0.20.2-core.jar
- Copy $MAPR_HOME/hadoop/hadoop-0.20.2/lib/hadoop-0.20.2-dev-core.jar into $PRD_HOME/lib
- Copy $MAPR_HOME/hadoop/hadoop-0.20.2/lib/maprfs-0.1.jar into $PRD_HOME/lib
- Linux x64:
- Add "-Djava.library.path=/opt/mapr/hadoop/hadoop-0.20.2/lib/native/Linux-amd64-64" to the last line in $PRD_HOME/report-designer.sh
- For MacOS:
- Add "-Djava.library.path=/opt/mapr/hadoop/hadoop-0.20.2/lib/native/Mac_OS_X-x86_64-64" to the "VMOptions" entry in $PRD_HOME/Pentaho\ Report\ Designer.app/Contents/Info.plist
Hadoop Node Configuration
Download the Pentaho Hadoop Node Distribution (PHD):
Ubuntu: phd-ce-mapr-bigdata-preview-4.3_all.deb
RedHat/CentOS: phd-ce-mapr-bigdata-preview-4.3.noarch.rpm
All TaskTracker nodes must have the pentaho-mapreduce (PHD) package installed on them. Our packages require the MapR TaskTracker (mapr-tasktracker
) package being installed.
From a high level the packages perform the following steps:
- Extract the Pentaho Hadoop Node Distribution archive into /opt/pentaho/pentaho-mapreduce
- Update $HADOOP_HOME/conf/hadoop-env.sh and add /opt/pentaho/pentaho-mapreduce/lib/* to the HADOOP_CLASSPATH
- Update $HADOOP_HOME/conf/mapred-site.xml and add the following properties if they do not exist:
<property> <name>pentaho.kettle.home</name> <value>/opt/pentaho/pentaho-mapreduce</value> </property> <property> <name>pentaho.kettle.plugins.dir</name> <value>/opt/pentaho/pentaho-mapreduce/plugins</value> </property>
RedHat/CentOS
Install
rpm -i phd-ce-mapr-bigdata-preview-4.3.noarch.rpm
Upgrade
rpm -U --force phd-ce-mapr-bigdata-preview-4.3.noarch.rpm
Ubuntu
Install
dpkg -i phd-ce-mapr-bigdata-preview-4.3_all.deb
Upgrade
Remove then reinstall:
dpkg -r pentaho-mapreduce
dpkg -i phd-ce-mapr-bigdata-preview-4.3_all.deb
Restart JobTracker and TaskTracker
To complete the installation you need to restart the JobTracker and TaskTracker nodes so the HADOOP_CLASSPATH changes take effect.