Job Templates
Section Contents
Job Templates#
Job templates allow you to run multiple job instances with different runtime parameter values from a single template definition. The SDK allows interaction with job templates including template creation, starting a job from a template, spawning multiple job instances from a template, and template deletion.
Creating a Job Template#
To create a new job template and add it to the Platform, use the streamsets.sdk.sch_models.JobBuilder
class. Use the streamsets.sdk.ControlHub.get_job_builder()
method to instantiate the builder object.
Simply pass job_template=True
to the streamsets.sdk.sch_models.JobBuilder.build()
method, and then pass the
resulting job object to the streamsets.sdk.ControlHub.add_job()
method:
job_builder = sch.get_job_builder()
job_template = job_builder.build('Job Template using SDK',
pipeline=simple_pipeline,
job_template=True,
runtime_parameters={'x': 'y', 'a': 'b'})
sch.add_job(job_template)
Starting Job Instances Using a Job Template#
Starting a job from a job template is very similar to starting any other job - with a few additional
requirements. Due to the nature of a job template, it expects that various values will be supplied for the
runtime_parameters
at start time, and likewise expects an instance_name_suffix
to be provided at start time.
The runtime_parameters
is optional and will default to the job template’s default set of parameters if none are
supplied at start time.
The instance_name_suffix
is appended to the end of the job template name to uniquely identify each job instance
started from the template. Additional details on the suffix can be found in the StreamSets Platform documentation,
and in the API reference for the streamsets.sdk.ControlHub.start_job_template()
method.
To start a job instance from a template, use the streamsets.sdk.ControlHub.start_job_template()
method and pass in the runtime_parameters
and instance_name_suffix
values:
# Get the job template to start the job from
job_template = sch.jobs.get(job_name='Job Template using SDK')
# Generate some runtime parameters to be used in the job
runtime_parameters = [{'x': '1', 'a': 'b'}, {'x': '2', 'a': 'b'}]
# Pass in the runtime parameters, specify the suffix type as 'PARAM_VALUE' and the corresponding parameter_name
# as 'x', i.e. the value 'x' will be appended to the end of this job's name upon startup
jobs = sch.start_job_template(job_template,
instance_name_suffix='PARAM_VALUE',
parameter_name='x',
runtime_parameters=runtime_parameters)
Spawning Multiple Job Instances Using the Same Parameters#
The SDK also allows you to spawn multiple job instances from the same template, all using the same runtime parameters.
Simply include a value for number_of_instances
in the streamsets.sdk.ControlHub.start_job_template()
method
to specify the number of Job instances to be spawned using the specified runtime parameters:
job_template = sch.jobs.get(job_name='Job Template using SDK')
jobs = sch.start_job_template(job_template, number_of_instances=3)
In this case, since runtime_parameters
is not specified, the default set of parameters specified when creating the
Job Template are used.
Deleting a Job Template#
Deleting a job template is identical to the steps required to delete a job from the Platform. Simply retrieve
the streamsets.sdk.sch_models.Job
instance you wish to delete, and pass it to the streamsets.sdk.ControlHub.delete_job()
method:
sch.delete_job(job_template)