A collection of useful Hadoop tools.

To use the tools build and use the file target/hadoop-tools-*-job.jar.

$ hadoop jar target/hadoop-tools-0.0.1-SNAPSHOT-job.jar [-command] [options]

Current commands are

-blockFinder

This tool will find the blocks that make up the file(s) specified and report
their names and locations. I found this useful once to identify the host that
was generating some files in HDFS. Using the information from this tool and
searching the logs of one of the nodes reported I was able to find the address
of the machine generating the file.


-fixReplication

This tool is not yet complete and will do nothing useful. Currently it should
report any files that are open and any that have blocks that violate the block
placement policy.


If you have any questions (or patches) contact jamesg@last.fm