Skip to content
/ oddjob Public

useful JVM classes for the mrjob hadoop streaming framework

Notifications You must be signed in to change notification settings

jblomo/oddjob

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

oddjob

This project gathers useful JVM classes for the mrjob hadoop streaming framework. Use the uberjar built by this project as an arguement to hadoop-streaming's -libjars option to use the classes below.

See also:

https://github.com/Yelp/mrjob
https://github.com/klbostee/feathers

Classes

oddjob.MultipleTextOutputFormatByKey - Writes to the directories specified by the key

The key of your job output will be used as the file path. Both the key and the value will be written to the resulting tab delimited text files.

eg rows:
filename1 {"some values": "other JSON"}
otherfile {"other values": "more JSON"}

in [outputdir]/filename1/part-00000 will be written filename1 {"some values": "other JSON"}
in [outputdir]/otherfile/part-00000 will be written otherfile {"other values": "more JSON"}

oddjob.MultipleValueOutputFormat - Writes to the directories specified by the key, and only writes the value

The key of your job output will be used as the file path. Only the value will be written to the resulting files.

eg rows:
filename1 {"some values": "other JSON"}
otherfile {"other values": "more JSON"}

in [outputdir]/filename1/part-00000 will be written {"some values": "other JSON"}
in [outputdir]/otherfile/part-00000 will be written {"other values": "more JSON"}

oddjob.MultipleLeafValueOutputFormat - Writes to the file specified by the key, and only writes the value

WARNING: advanced usage only. You probably want MultipleValueOutputFormat instead. This class allows the developer to specify the entire path of the output file, including the "leaf" file (usually 'part-00XXX'). This means the job /must/ output unique keys per process. For example, you may use this format if you guaruntee each key is unique per reducer.

eg rows:
full/filename {"some values": "other JSON"}
otherpath/filename {"other values": "more JSON"}

in [outputdir]/full/filename will be written {"some values": "other JSON"}
in [outputdir]/otherpath/filename will be written {"other values": "more JSON"}

oddjob.MultipleCSVOutputFormat - Writes to the directories specified by the first element in a row

The output key of your job must be a comma separated row, fields optionally enclosed by double quotes. The first element will be used as the subdirectory, and the written row will not include that first element.

eg rows:
"even",16,4
"odd",25,5

in [outputdir]/even/part-00000 will be written 16,4
in [outputdir]/odd/part-00000 will be written 25,5

oddjob.MultipleJSONOutputFormat - Writes to the directories specified by the first element in the key

The output key of your job must be a JSON formatted array. The first element will be used as the subdirectory, and the second element will be used for key written to the file.

eg rows:
[17, "realkey"] {"some values": "other JSON"}
[22, "other realkey"] {"other values": "more JSON"}

in [outputdir]/17/part-00000 will be written "realkey" {"some values": "other JSON"}
in [outputdir]/22/part-00000 will be written "other realkey" {"other values": "more JSON"}

oddjob.ManifestTextInputFormat - Manifest files are inputs with a list of paths to use as the real input.

Paths may be directories, globs, or files and will be expanded appropriately. Unlike most InputFormats, this class will silently ignore missing and unmatched paths in the manifest file.

Manifest files are useful for getting around the limit imposed by Amazon EMR on length of the input paths.

Usage

../lein uberjar
python examples/mr_wc.py --verbose -r emr  --hadoop-version 0.20  --bootstrap-file oddjob-*-standalone.jar  --bootstrap-cmd 'cp oddjob-*-standalone.jar /home/hadoop/' --hadoop-arg -libjars --hadoop-arg oddjob-*-standalone.jar input.txt

This version of oddjob is designed to be used as a -libjar argument with hadoop 0.20. -libjar arguments must be available to the hadoop script when you run the job, so you can use an inline bootstrap-cmd as shown, set it up in your config, or have an AMI with it already installed.

License

Copyright 2010-2012 Yelp

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

About

useful JVM classes for the mrjob hadoop streaming framework

Resources

Stars

Watchers

Forks

Packages

No packages published