def azkaban_mongo_job():
project_2 = Project('LeadDataStats')
project_2.add_job('UploadDataStats', Job({'type': 'command', 'command': 'python /home/msingh/Documents/'
'PycharmProjects/AzkabanTest/mongo/MongoDataPush.py'}, {'dependencies': 'MongoStart'}))
project_3 = Project('MongoDataUpload')
project_3.add_job('UploadDataStatus', Job({'type': 'command', 'command': 'echo "Data successfully uploaded"'},
{'dependencies': 'UploadDataStats'}))
def build_project(project_name, global_props, project_props, jobs, files,
version):
logger.info("Building workflow %s, version: %s.", project_name, version)
project = Project(project_name, root=os.curdir, version=version)
project.properties = global_props
project.properties.update(project_props)
for job_name, job_definition in jobs.items():
project.add_job(job_name, Job(job_definition))
for file, target in files:
project.add_file(file, target)
return project
},
'jvm.args.mapred': {
'max.split.size': 2684354560,
'min.split.size': 2684354560,
},
}
# list of pig job options
OPTIONS = [
{
'pig.script': 'first.pig'
},
{
'pig.script': 'second.pig',
'dependencies': 'first.pig'
},
{
'pig.script': 'third.pig',
'param': {
'foo': 48
}
},
{
'pig.script': 'fourth.pig',
'dependencies': 'second.pig,third.pig'
},
]
for option in OPTIONS:
PROJECT.add_job(option['pig.script'], PigJob(DEFAULTS, option))
#!/usr/bin/env python
# encoding: utf-8
"""Simple Azkaban project configuration script."""
from azkaban import Job, Project
project = Project('foo')
project.add_job('bar', Job({'type': 'command', 'command': 'echo "hi!"'}))
if __name__ == '__main__':
project.main()
'basic_flow': Job({'type': 'noop' , 'dependencies': 'basic_step_5.cmd,basic_step_6.cmd,basic_step_7.cmd,basic_step_8.cmd'}),
# `template_flow` example
# • Demonstrates using one flow as a "template" that is embedded in another flow and reused multiple times.
# • The only work performed by job in this example template is to echo out the variables it receives to the log.
# NOTE: We have to `chmod 777` our script to make sure Azkaban can run it.
'_template_chmod.cmd': Job({'type': 'command', 'command': 'chmod 777 _echo.sh'}),
'_template_echo_1.cmd': Job({'type': 'command', 'command': './_echo.sh "echo_1" ${project_1} ${custom_1} ${custom_2}', 'dependencies': '_template_chmod.cmd'}),
'_template_echo_2.cmd': Job({'type': 'command', 'command': './_echo.sh "echo_2" ${project_1} ${custom_1} ${custom_2}', 'dependencies': '_template_echo_1.cmd'}),
'_template': Job({'type': 'noop' , 'dependencies': '_template_echo_2.cmd'}),
# • Each of the following subflows embeds *ALL* of the steps from `_template` using the `flow.name` key.
# • Each defines `custom_1` and `custom_2` keys which are passed as variables ${custom_1} and ${custom_2} to `_template` during execution.
'start.noop': Job({'type': 'noop'}),
'subflow_1.flow': Job({'type': 'flow', 'flow.name': '_template', 'dependencies': 'start.noop', 'custom_1': 'subflow1-val1', 'custom_2': 'subflow1-val2'}),
'subflow_2.flow': Job({'type': 'flow', 'flow.name': '_template', 'dependencies': 'start.noop', 'custom_1': 'subflow2-val1', 'custom_2': 'subflow2-val2'}),
'subflow_3.flow': Job({'type': 'flow', 'flow.name': '_template', 'dependencies': 'start.noop', 'custom_1': 'subflow3-val1', 'custom_2': 'subflow3-val2'}),
'subflow_4.flow': Job({'type': 'flow', 'flow.name': '_template', 'dependencies': 'start.noop', 'custom_1': 'subflow4-val1', 'custom_2': 'subflow4-val2'}),
'workflow': Job({'type': 'noop', 'dependencies': 'subflow_1.flow,subflow_2.flow,subflow_3.flow,subflow_4.flow'})
}
for name, job in JOBS.items():
PROJECT.add_job(name, job)
# The CLI requires any non-job files to be explicitly included.
# Must declare the `root` in the project in order for this to work.
FILES = {
'./_echo.sh': '_echo.sh'
}
for file, name in FILES.items():
PROJECT.add_file(file, name)
PROJECT = Project('azkabancli_sample', root=__file__)
PROJECT.properties = {
'user.to.proxy': 'production_user',
'hdfs.root': '/jobs/sample/'
}
# dictionary of jobs, keyed by job name
JOBS = {
'gather_data':
Job({
'type': 'hadoopJava',
'job.class': 'sample.GatherData',
'path.output': '${hdfs.root}data.avro', # note the property use here
}),
# ...
}
for name, job in JOBS.items():
PROJECT.add_job(name, job)
# Test project
# ------------
#
# This project is an exact copy of the production project which can be used
# to debug / test new features independently from the production flows.
TEST_PROJECT = Project('sample_test', root=__file__)
TEST_PROJECT.properties = {'user.to.proxy': getuser(), 'hdfs.root': 'sample/'}
PROJECT.merge_into(TEST_PROJECT)
from azkaban import Job, Project
project = Project('foo')
project.add_file('./jobs.py', 'jobs.py')
project.add_job('bar', Job({'type': 'command', 'command': 'cat jobs.py'}))
project = Project('foo', root=__file__)
defaults = {
'user.to.proxy': getuser(),
'mapred': {
'max.split.size': 2684354560,
'min.split.size': 2684354560,
},
}
project.add_job(
'first_pig_script',
PigJob(
'path/to/first_script.pig', # assume it exists
defaults,
)
)
project.add_job(
'second_pig_script',
PigJob(
'path/to/second_script.pig', # assume it also exists
defaults,
{'mapred.job.queue.name': 'special'},
)
)
project.add_job(
'final_job',
from azkaban import PigJob, Project
from getpass import getuser
PROJECT = Project('azkabancli_sample', root=__file__)
# default options for all jobs
DEFAULTS = {
'user.to.proxy': getuser(),
'param': {
'input_root': 'sample_dir/',
'n_reducers': 20,
},
'jvm.args.mapred': {
'max.split.size': 2684354560,
'min.split.size': 2684354560,
},
}
# list of pig job options
OPTIONS = [
{'pig.script': 'first.pig'},
{'pig.script': 'second.pig', 'dependencies': 'first.pig'},
{'pig.script': 'third.pig', 'param': {'foo': 48}},
{'pig.script': 'fourth.pig', 'dependencies': 'second.pig,third.pig'},
]
for option in OPTIONS:
PROJECT.add_job(option['pig.script'], PigJob(DEFAULTS, option))
