Doing Jobs Dinamically with Azkaban
Project description
Auror Core
Simple Flow creation for Azkaban
Install
pip install auror_core
Supported Job Types
V1
- Command
- Flow
V2
- Command
Usage
- Creating a simple Azkaban flow with one command
- Creating a simple V2 Azkaban flow with one command
- Creating Flows with dependencies
- Sharing job attributes
- Job with extra customization and configuration
- Integrating with Flow (just for V1)
- Using Flow Params
- Using Flow Environment Variables
- Using Flow Environment Variables and Params
- Join multiple variables in one
- Load jobs from YAML File (just for V2)
- Dump memory flows to a Python File (just for V2)
Creating a simple Azkaban flow with one command
You just need to import job type and project
from auror_core.v1.job import Job, Command
from auror_core import Project
com1 = Job()\
.as_type(Command)\
.with_name("commands job")\
.with_command("bash echo 1")\
.with_another_command("bash echo 2")
Project("folder_to_generate_files", com1).write()
Creating a simple V2 Azkaban flow with one command
V2 flow is implemented under v2 subfolder with same job types
from auror_core.v2.job import Job, Command
from auror_core import Project
com1 = Job()\
.as_type(Command)\
.with_name("commands job")\
.with_command("bash echo 1")\
.with_another_command("bash echo 2")
Project("folder_to_generate_files", com1).is_v2().write()
Creating Flows with dependencies
from auror_core.v2.job import Job, Command
from auror_core import Project
com1 = Job()\
.as_type(Command)\
.with_name("commands job")\
.with_command("bash echo 1")\
.with_another_command("bash echo 2")
com2 = Command()\
.with_name("sub command job")\
.with_command("bash echo 1")\
.with_dependencies(com1)
Project("folder_to_generate_files", com1, com2).is_v2().write()
Sharing job attributes
Organize jobs with same configuration
from auror_core.v2.job import Command
from auror_core import Project
com = Command()\
.with_command("bash echo 1")
com1 = com.with_name("commands job")\
.with_another_command("bash echo 2")
com2 = com.with_name("sub command job")\
.with_dependencies(com1)
Project("folder_to_generate_files", com1, com2).is_v2().write()
Job with extra customization and configuration
Simulating a Command with base Job (NOT RECOMMENDED)
from auror_core.v1.job import Job
from auror_core import Project
com1 = Job()\
.with_name("commands job")\
.with_(command="bash echo 1")
com1._type = "command"
Project("folder_to_generate_files", com1).write()
Integrating with Flow (just for V1)
V2 already have flow included
from auror_core.v1.job import Command, Flow, Job
from auror_core import Project
com1 = Command()\
.with_name("commands job")\
.with_command("bash echo 1")
flow = Job()\
.as_type(Flow)\
.with_name("flow")\
.with_dependencies(com1)
Project("folder_to_generate_files", com1, flow).write()
Using Flow Params
from auror_core.v2.job import Command
from auror_core.v2.params import Params
from auror_core import Project
params = Params(
teste1="my test",
variable="my variable"
)
com = Command()\
.with_command("bash echo ${variable}")
Project("folder_to_generate_files", com)\
.is_v2()\
.with_params(params)\
.write()
Using Flow Environment Variables
from auror_core.v2.job import Command
from auror_core.v2.params import Env
from auror_core import Project
env = Env(
TESTE="my test",
VARIABLE="my variable"
)
com = Command()\
.with_command("bash echo $VARIABLE")
Project("folder_to_generate_files", com)\
.is_v2()\
.with_params(env)\
.write()
Using Flow Environment Variables and Params
from auror_core.v2.job import Command
from auror_core.v2.params import Env, Params
from auror_core import Project
env = Env(
TESTE="my test",
VARIABLE="my variable"
)
params = Params(
teste1="my test",
variable="my variable"
)
com = Command()\
.with_command("bash echo $VARIABLE ${variable}")
Project("folder_to_generate_files", com)\
.is_v2()\
.with_params(params, env)\
.write()
Join multiple variables in one
from auror_core.v2.job import Command
from auror_core.v2.params import Env
from auror_core import Project
env = Env(
TESTE="env test",
VARIABLE="env variable"
)
params = Params(
teste1="my test",
variable="my variable"
)
one_param = ParamsJoin("params_strange_name", ",") ## param name and separator
com = Command()\
.with_command("bash echo ${params_strange_name}")
## it will print: my test,my variable,env test,env variable
## THERE IS NO ORDER GUARANTEE, JUST Python 3.6 >
Project("folder_to_generate_files", com)\
.is_v2()\
.with_params(one_param(params, env))\
.write()
Load jobs from YAML File (just for V2)
You can find some YAML File examples on Azkaban Flow Documentation
from auror_core.v2.loader import Loader
loader = Loader('/path/to/file/flow.yaml')
jobs = loader.as_job_objects()
Or you can export these jobs as a Python File
from auror_core.v2.loader import Loader
loader = Loader('/path/to/file/flow.yaml')
jobs = loader.as_python_file('/path/to/desired/directory') # will be dumped with 'flow.py' name
Dump memory flows to a Python File (just for V2)
from auror_core.v2.dumper import Dumper
com1 = Job() \
.with_name("commands job 1") \
.with_(command="bash echo 1")
com2 = Job()\
.with_name("commands job 2")\
.with_(command="bash echo 2")
dumper = Dumper('/path/to/desired/directory') # will be dumped with 'flow.py' name
dumper.dump_jobs(com1, com2)
Plugins
Plugins are just extensions from auror_core
There is a cookiecutter for new azkaban jobtypes with Auror template too: https://github.com/globocom/azkaban-jobtype-cookiecutter
We already have email plugin: https://github.com/globocom/azkaban-jobtype-email
Contribute
For development and contributing, please follow Contributing Guide and ALWAYS respect the Code of Conduct
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.