Ali Kanso

I see what you mean. If we use a scheduling framework like volcano, we would introduce a dependency on the Volcano api: `apiVersion: scheduling.volcano.sh` Moreover, part of deploying KubeRay would...

yes good idea. I guess we would also need to specify the CR instance name for the raycluster. kubectl ray scale `$raycluster` $groupName $replica

I guess scaling down might require an optional flag, which is the name of the pod to scale down. Because in certain situations we cannot take down a random pod,...

one thing to note here, is that multiple KubeRay instances do **not** improve performance, on the contrary, it might slow it down (due to leader election failure). Since at any...

I agree, since the operator affects the K8s cluster state, we **do not want two active operator pods running at the same time.** Since the startup time is negligible for...

can we change the **PR name**, from `annotate` to enrich cluster status with head pod ip?