You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, if we use the with statement to run a pipeline, it waits until finish by default with no way to configure it. The only way to run without waiting is by calling pipeline.run() explicitly without the with statement context. This is not consistent and users now have to be aware of two different ways of calling.
Ideally, we would like to define pipelines with the with statement always to have a "single way to do things". Running without waiting is useful, for example, to run a pipeline in Cloud Build, Cloud Run, templates, or other serverless frameworks.
Right now the only way to do it is by changing the code to remove the with statement and calling pipeline.run() manually, which might not be always possible. With a pipeline option we can just pass that from the command line arguments, or even when testing.
I'm opening this for Python since many Python samples would greatly benefit from this, but I can see this being useful in Java and Go as well.
Imported from Jira BEAM-12455. Original Jira may contain additional context.
Reported by: dcavazos.
The text was updated successfully, but these errors were encountered:
Currently, if we use the
with
statement to run a pipeline, it waits until finish by default with no way to configure it. The only way to run without waiting is by callingpipeline.run()
explicitly without thewith
statement context. This is not consistent and users now have to be aware of two different ways of calling.Ideally, we would like to define pipelines with the
with
statement always to have a "single way to do things". Running without waiting is useful, for example, to run a pipeline in Cloud Build, Cloud Run, templates, or other serverless frameworks.Right now the only way to do it is by changing the code to remove the
with
statement and callingpipeline.run()
manually, which might not be always possible. With a pipeline option we can just pass that from the command line arguments, or even when testing.I'm opening this for Python since many Python samples would greatly benefit from this, but I can see this being useful in Java and Go as well.
Imported from Jira BEAM-12455. Original Jira may contain additional context.
Reported by: dcavazos.
The text was updated successfully, but these errors were encountered: