Hi Debasish Ghosh, Thanks for the attention on native K8s integration of Flink.
1. For volumes and volumes mount, it is not supported now. And we are trying to get it done via pod template. Refer here[1] for more information. 2. Currently, on different deployments, Flink has different cpu config options. But for the memory, all the deployments share the same config options. You could find more information here[2]. * yarn.appmaster.vcores * yarn.containers.vcores * kubernetes.jobmanager.cpu * kubernetes.taskmanager.cpu 3. You are right. This class is PublicEvolving and we may introduce more config options in the future(e.g. pod template related). [1]. https://lists.apache.org/thread.html/rf2e7b9be96f2bd5106d08ffb573d55f70a8acfb0b814a21d8b50d747%40%3Cdev.flink.apache.org%3E [2]. https://ci.apache.org/projects/flink/flink-docs-master/deployment/memory/mem_setup.html Debasish Ghosh <ghosh.debas...@gmail.com> 于2020年12月21日周一 下午3:21写道： > Hello - > > In > https://github.com/apache/flink/blob/master/flink-kubernetes/src/main/java/org/apache/flink/kubernetes/configuration/KubernetesConfigOptions.java > the various supported options are declared as constants. > > I see that there is no support for options like Volumes and VolumeMounts. > Also I see entries for JOB_MANANGER_CPU and TASK_MANAGER_CPU but not for > JOB_MANAGER_MEMORY and TASK_MANAGER_MEMORY. How do we accommodate these if > we want to pass them as well ? I see that the class is annotated > with @PublicEvolving - just wanted to clarify if these are planned to be > added in future. > > regards. > -- > Debasish Ghosh > http://manning.com/ghosh2 > http://manning.com/ghosh > > Twttr: @debasishg > Blog: http://debasishg.blogspot.com > Code: http://github.com/debasishg >
