Yashpal comments

Results 9 comments of


                                            Yashpal

volume is in pendig state in case of waitforfirstconsumer

@w3aman It may or may not be scheduled on node-2 in above example. See it's not guaranteed that pod will be scheduled on the node having free storage space since...

expose pvc/pod level metrics (if not already exposed by kubelet and cadvisor)

Thanks @kmova. Let me know as the dashboard gets ready & pushed somewhere. I would like to try them out in our playground clusters.

NodeLifeCycleController: all pods are marked as not ready making workload services unavailable during network partition b/w master & worker nodes

/sig api-machinery

NodeLifeCycleController: all pods are marked as not ready making workload services unavailable during network partition b/w master & worker nodes

Thanks @ehashman for the response. > If you have a network partition that completely severs the control plane from worker nodes, it makes sense to me that the node lifecycle...

NodeLifeCycleController: all pods are marked as not ready making workload services unavailable during network partition b/w master & worker nodes

/reopen /remove-lifecycle rotten

NodeLifeCycleController: all pods are marked as not ready making workload services unavailable during network partition b/w master & worker nodes

/reopen /remove-lifecycle rotten

Feature Request - Allow Configuring Volume Limits Per Node

/reopen /remove-lifecycle rotten

Feature Request - Allow Configuring Volume Limits Per Node

/reopen

coredns doesn't perform better despite having more cores

yes @lobshunter that is correct. I think [lwn article](https://lwn.net/Articles/542629/) explain the improvements and few caveats (esp. with TCP) of using `SO_REUSEPORT` option. Last week, I had validated the improvements by...