I need to run a Flink job in production, I have just one machine (is not a cluster). I'm trying to setup my Flink job for production and do it in the best way possible, different as I have it right now (running as a java application java -jar flink_job.jar). I was reading in https://ci.apache.org/projects/flink/flink-docs-stable/ops/deployment/#deployment-targets that this deployment mode is only for testing or experimentations, so, for production I think that I should move to standalone cluster, but, in Standalone cluster deployment mode says that I need to setup one master and at least one worker. My question, that I have to run in production just for one node (master):
- What is the best way to run flink in production on my case?
- Can I have my master and worker in the same JVM as the same way that Local cluster does?
Please can someone clarify a little bit this for me? Thanks. Kind regards.
