Stocator and Apache Hadoop

Stocator can be used natively with Hadoop MapReduce

• Download Hadoop 2.7.x

• Add stocator-all-in-one.jar to the HADOOP_CLASSPATH

• Configure core-sites.xml to use Stocator

• Certain Hadoop flows requires default fs, which is file:///. To over come this edit core-sites.xml and add

  <property> <name>fs.default.name</name> <value>swift2d://container.service/test</value> </property>

alternatively you can provide this parameter as a command line argument. For example

./bin/hadoop jar hadoop-mapreduce-client-jobclient-2.7.2-tests.jar TestDFSIO -Dtest.build.data=swift2d://container.service/data -Dfs.default.name=swift2d://container.service/ -write -nrFiles 3 -fileSize 10MB -resFile /local/out/summary-write.txt