Hadoop I/O - Hadoop: The Definitive Guide - page 128

Each part of the final output is compressed; in this case, there is a single part:

% gunzip -c output/part-r-00000.gz
1949 111
1950 22

If you are emitting sequence files for your output, you can set the mapre-
duce.output.fileoutputformat.compress.type property to control the
type of compression to use. The default is RECORD , which compresses individual records.
Changing this to BLOCK , which compresses groups of records, is recommended because it
compresses better (see The SequenceFile format ) .

There is also a static convenience method on SequenceFileOutputFormat called
setOutputCompressionType() to set this property.
The configuration properties to set compression for MapReduce job outputs are summar-
ized in Table 5-5 . If your MapReduce driver uses the Tool interface (described in Gener-
icOptionsParser, Tool, and ToolRunner ), you can pass any of these properties to the pro-
gram on the command line, which may be more convenient than modifying your program
to hardcode the compression properties.

Table 5-5. MapReduce compression properties

Property name Type Default value

mapreduce.output.fileoutputformat.compress boolean false

mapreduce.output.fileoutputformat.compress.codec Class
name

org.apache.hadoop.io.compress.DefaultCodec The com-

mapreduce.output.fileoutputformat.compress.type String RECORD