mingzhang-ybps

How to support scheduling custom resource without "replicas" characteristic among managed clusters

2

**Please provide an in-depth description of the question you have**: Currently, Karmada provides flexiable scheduling against normal kubernetes workloads such as deployment, job，etc. From the implemention https://github.com/karmada-io/karmada/blob/master/pkg/scheduler/core/generic_scheduler.go#L82, the generic scheduler...

Reduce overhead of work objects for numerous member clusters scenario

2

**Please provide an in-depth description of the question you have**: Karmada create a work object in its control plane mapping to a workload which will be distributed into member cluster....

[Bug] Shutdown node and wait some minutes to start it again, the pod stucks in Terminating state

2

**Steps to reproduce** 1. create a deployment which uses csi-s3 to provision filesystem PV  2. then shutdown the node ltaucenw8ridob54 which hosts the first pod nginx-deployment-7d94b7b565-2rrqr 3. after a...

Fix issue https://github.com/yandex-cloud/k8s-csi-s3/issues/149