We don't have a recommendation, this will depend on the number of jobs you manage.
If you are managing a few jobs let's say < 50, the default resource configuration probably works well. The operator reports JVM metrics similar to Flink jobs so if you see longer GC pauses you could simply add more memory/cpu as you scale the number of managed jobs. The operator is not very CPU intensive, but it uses some heap space for caching information about the jobs. Gyula On Mon, Nov 28, 2022 at 8:01 AM hjw <18814122...@163.com> wrote: > Flink Kubernetes Operator is a server is responsible to reconcile the > FlinkDeployment. > The operator will continuously monitor the status of Flink jobs.The > stability of the operator's service is also important. > What is the recommended cpu.memory configuration of the Flink kubernetes > operator in the production environment ? > > > -- > Best， > Hjw >
