Live migration of Spark Streaming job from one configuration to another
I'm a PhD student and I'm trying to model the performances (processing delay, throughput) of Spark Streaming jobs, and I wonder whether there is a way to do live migration of a Spark Streaming job from one configuration to another?
(i.e. without having to interrupt the job and then re-submit using the sparkSubmit command with the new configuration parameters).
Is it possible in both APIs (DStreams and Structured Streaming)?