Which of the following is NOT a role of the Kube-Scheduler?

Which of the following is NOT a role of the Kube-Scheduler?

• A. It determines which nodes are valid placements for Pods according to resources and constraints
• B. It updates the desired state of the cluster
• C. It assigns Pods to nodes
• D. It respects resource requirements and constraints

Answer: (B)

The scheduler’s job is to pick a suitable node for a Pod and bind it there, based on the Pod’s resource requests and the cluster’s constraints. It evaluates which nodes have enough CPU and memory, honors taints and tolerations, node selectors, and any scheduling policies, then selects a fit and binds the Pod to that node. That binding action is what effectively assigns the Pod to a node. Crucially, the scheduler does not update the desired state of the cluster. Updating the desired state—such as how many replicas should be running or how a workload should be configured—is the responsibility of the control plane controllers (for example, the Deployment or ReplicaSet controllers) and the API server, which modify the persistent state and allow the scheduler to reconcile actual state to the desired state. So the scheduler’s roles include determining valid placements on nodes, assigning (binding) Pods to nodes, and ensuring the Pod’s resource requirements and constraints are respected. Updating the desired state is outside its responsibilities.

The scheduler’s job is to pick a suitable node for a Pod and bind it there, based on the Pod’s resource requests and the cluster’s constraints. It evaluates which nodes have enough CPU and memory, honors taints and tolerations, node selectors, and any scheduling policies, then selects a fit and binds the Pod to that node. That binding action is what effectively assigns the Pod to a node.

Crucially, the scheduler does not update the desired state of the cluster. Updating the desired state—such as how many replicas should be running or how a workload should be configured—is the responsibility of the control plane controllers (for example, the Deployment or ReplicaSet controllers) and the API server, which modify the persistent state and allow the scheduler to reconcile actual state to the desired state.

So the scheduler’s roles include determining valid placements on nodes, assigning (binding) Pods to nodes, and ensuring the Pod’s resource requirements and constraints are respected. Updating the desired state is outside its responsibilities.