blob: 15dfe8897f8dd1591568b16f7c175ba722a29fb4 [file] [log] [blame]
# Copyright (c) 2010 Spotify AB
# Copyright (c) 2010-2011 Yelp
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the
# "Software"), to deal in the Software without restriction, including
# without limitation the rights to use, copy, modify, merge, publish, dis-
# tribute, sublicense, and/or sell copies of the Software, and to permit
# persons to whom the Software is furnished to do so, subject to the fol-
# lowing conditions:
#
# The above copyright notice and this permission notice shall be included
# in all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABIL-
# ITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT
# SHALL THE AUTHOR BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
# WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
class Step(object):
"""
Jobflow Step base class
"""
def jar(self):
"""
:rtype: str
:return: URI to the jar
"""
raise NotImplemented()
def args(self):
"""
:rtype: list(str)
:return: List of arguments for the step
"""
raise NotImplemented()
def main_class(self):
"""
:rtype: str
:return: The main class name
"""
raise NotImplemented()
class JarStep(Step):
"""
Custom jar step
"""
def __init__(self, name, jar, main_class=None,
action_on_failure='TERMINATE_JOB_FLOW', step_args=None):
"""
A elastic mapreduce step that executes a jar
:type name: str
:param name: The name of the step
:type jar: str
:param jar: S3 URI to the Jar file
:type main_class: str
:param main_class: The class to execute in the jar
:type action_on_failure: str
:param action_on_failure: An action, defined in the EMR docs to take on failure.
:type step_args: list(str)
:param step_args: A list of arguments to pass to the step
"""
self.name = name
self._jar = jar
self._main_class = main_class
self.action_on_failure = action_on_failure
if isinstance(step_args, basestring):
step_args = [step_args]
self.step_args = step_args
def jar(self):
return self._jar
def args(self):
args = []
if self.step_args:
args.extend(self.step_args)
return args
def main_class(self):
return self._main_class
class StreamingStep(Step):
"""
Hadoop streaming step
"""
def __init__(self, name, mapper, reducer=None, combiner=None,
action_on_failure='TERMINATE_JOB_FLOW',
cache_files=None, cache_archives=None,
step_args=None, input=None, output=None,
jar='/home/hadoop/contrib/streaming/hadoop-streaming.jar'):
"""
A hadoop streaming elastic mapreduce step
:type name: str
:param name: The name of the step
:type mapper: str
:param mapper: The mapper URI
:type reducer: str
:param reducer: The reducer URI
:type combiner: str
:param combiner: The combiner URI. Only works for Hadoop 0.20 and later!
:type action_on_failure: str
:param action_on_failure: An action, defined in the EMR docs to take on failure.
:type cache_files: list(str)
:param cache_files: A list of cache files to be bundled with the job
:type cache_archives: list(str)
:param cache_archives: A list of jar archives to be bundled with the job
:type step_args: list(str)
:param step_args: A list of arguments to pass to the step
:type input: str or a list of str
:param input: The input uri
:type output: str
:param output: The output uri
:type jar: str
:param jar: The hadoop streaming jar. This can be either a local path on the master node, or an s3:// URI.
"""
self.name = name
self.mapper = mapper
self.reducer = reducer
self.combiner = combiner
self.action_on_failure = action_on_failure
self.cache_files = cache_files
self.cache_archives = cache_archives
self.input = input
self.output = output
self._jar = jar
if isinstance(step_args, basestring):
step_args = [step_args]
self.step_args = step_args
def jar(self):
return self._jar
def main_class(self):
return None
def args(self):
args = []
# put extra args BEFORE -mapper and -reducer so that e.g. -libjar
# will work
if self.step_args:
args.extend(self.step_args)
args.extend(['-mapper', self.mapper])
if self.combiner:
args.extend(['-combiner', self.combiner])
if self.reducer:
args.extend(['-reducer', self.reducer])
else:
args.extend(['-jobconf', 'mapred.reduce.tasks=0'])
if self.input:
if isinstance(self.input, list):
for input in self.input:
args.extend(('-input', input))
else:
args.extend(('-input', self.input))
if self.output:
args.extend(('-output', self.output))
if self.cache_files:
for cache_file in self.cache_files:
args.extend(('-cacheFile', cache_file))
if self.cache_archives:
for cache_archive in self.cache_archives:
args.extend(('-cacheArchive', cache_archive))
return args
def __repr__(self):
return '%s.%s(name=%r, mapper=%r, reducer=%r, action_on_failure=%r, cache_files=%r, cache_archives=%r, step_args=%r, input=%r, output=%r, jar=%r)' % (
self.__class__.__module__, self.__class__.__name__,
self.name, self.mapper, self.reducer, self.action_on_failure,
self.cache_files, self.cache_archives, self.step_args,
self.input, self.output, self._jar)