Ahmed Ezzat
Ahmed Ezzat
Currently, the operator has a hardcoded TCP protocol https://github.com/dask/dask-kubernetes/blob/92714da5785709726f85c4c6ec92451f5c23ad04/dask_kubernetes/operator/controller/controller.py#L155-L159
Currently, the operator retires workers using the HTTP or RPC APIs however those only control the connected dask workers, the operator should take into count dask's Kubernetes worker pods that...
Implement a way to set a cool-down period for adaptive scaling instead of the hardcoded https://github.com/dask/dask-kubernetes/blob/92714da5785709726f85c4c6ec92451f5c23ad04/dask_kubernetes/operator/controller/controller.py#L816 e.g. ```yaml apiVersion: kubernetes.dask.org/v1 kind: DaskCluster metadata: annotations: kubernetes.dask.org/cooldown-until-interval: "30s" name: dask-f3a0c12f namespace: default...
This PR fixes an overwrite bug in the `DEFAULT_CHART_OPTIONS` as it was shared and mutated between chart instances resulting in a wrong rendering. Before: After: