To run this, copy a mongod executable to this directory. (You can get a copy here) Then, run it with ./sbt run <args>, where args are
- databaseName - the name of the database you are dumping from
- shardName - the shard you are dumping
- inputDir - mongod directory to dump from
- hdfsPath - path to dump data to
- dbPort - any free port for mongod to use
- localTmpDir - local path for temporary data
ThriftBsonInputFormat can be used to read BSON files generated in this way from MapReduce jobs. It's configured with:
conf.setInputFormat(classOf[ThriftBsonInputFormat])
conf.set(ThriftBsonInputFormat.thriftClass, classOf[MyThriftClass].getName)