wangyang0918 commented on pull request #15501: URL: https://github.com/apache/flink/pull/15501#issuecomment-847629287
I have verified this PR with a minikube cluster. It works well. Thanks @yittg for the great work. However, I still find a suspicious log which shows up every 5 seconds. It might be related with https://github.com/fabric8io/kubernetes-client/issues/2651. This will greatly increase the pressure of ETCD and affect the whole K8s cluster stability. Could you please have a look? ``` 2021-05-25 07:04:45,580 DEBUG [pool-5-thread-1] io.fabric8.kubernetes.client.informers.cache.Reflector [] - Listing items (4) for resource class io.fabric8.kubernetes.api.model.ConfigMap v3656102 ``` I have another comment about FLINK-22006. We may need to update the related doc here[1]. Because all the jobs now share a same web socket connection, as well as concurrent request. For `FlinkKubeClientFactory#trySetMaxConcurrentRequest`, I think we could still have it until we bump the fabric8 Kubernetes client version. [1]. https://ci.apache.org/projects/flink/flink-docs-master/docs/deployment/resource-providers/native_kubernetes/#configuring-flink-on-kubernetes -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
