Drain Running Job
A 3-step recipe with Google Cloud Dataflow. We can help you get access.
Recipe · 3 steps
Drain Running Job
Confirm running, request drain, and poll until JOB_STATE_DRAINED.
- Read the job and confirm it is currently in the JOB_STATE_RUNNING state before requesting a drain.
- Update the job's requested state to JOB_STATE_DRAINING so it stops ingesting new data while finishing in-flight work.
- Poll the job until it reports the terminal JOB_STATE_DRAINED state, retrying while it is still draining.
What you'll need to have ready
- accessToken: Google Cloud OAuth 2.0 access token used as a Bearer credential.
- projectId: The Google Cloud project id that owns the job.
- location: The regional endpoint that contains the job (e.g. us-central1).
- jobId: The id of the running streaming job to drain.
For your developer
Arazzo workflow google-cloud-dataflow:drain-running-job, 3 operations: getLocationJob → updateLocationJob → getLocationJob. Specs on APIs.io.