We could monitor and update labels on nodes for when IDE images are pulled on nodes, then use perfered node affinity for scheduling. This will help when scaling to make sure IDEs will be scheduled on nodes that have the image pulled already if available.
https://kubernetes.io/docs/tasks/configure-pod-container/assign-pods-nodes-using-node-affinity/
We could monitor and update labels on nodes for when IDE images are pulled on nodes, then use perfered node affinity for scheduling. This will help when scaling to make sure IDEs will be scheduled on nodes that have the image pulled already if available.
https://kubernetes.io/docs/tasks/configure-pod-container/assign-pods-nodes-using-node-affinity/