Hey community,
Flink provides several execution modes. One of which is the Application
Mode
(https://ci.apache.org/projects/flink/flink-docs-stable/deployment/resource-providers/standalone/#application-mode),
which seems to be particularly suitable for deploying a self-contained
software component, such as an event-driven microservice, that needs to
apply stream processing techniques.
I'm wondering if it's possible to configure Beam's Flink runner to run
Flink in application mode? That is to deploy its own Flink instance
instead of submitting the job to an already running cluster (i.e. in
session mode)?
Thanks
Sören
