Requestscpu

Requestscpu refers to the CPU resource request specified for a container in Kubernetes. It is defined as part of a Pod’s resource requirements, under the containers[].resources.requests.cpu field. The value denotes the amount of CPU that the scheduler should guarantee to allocate to the container when scheduling the pod.

How it works: The Kubernetes scheduler uses the request value to determine whether a node has enough available CPU to run the pod. The sum of all CPU requests on a node cannot exceed the node’s total CPU capacity. Requests are also used for accounting and influence the pod’s QoS class. If a node becomes resource-constrained, pods may be throttled or evicted based on QoS and priority.

Units and interpretation: CPU requests are expressed in cores or millicores. Common forms include 0.5 for half a core or 500m for 0.5 CPU. A value of 1 or 1000m represents one full CPU. If both requests and limits are set, they can differ; if only a request is set, Kubernetes uses the request for scheduling, while the limit governs maximum usage if enforced.

Relationship to limits and QoS: Requests establish a guaranteed minimum, while limits cap maximum usage. If both requests and limits are equal, the pod tends toward the Guaranteed QoS class. If limits exceed requests, the pod is Burstable; if neither is set, the pod is BestEffort. CPU throttling is applied by the container runtime when limits are exceeded.

Example: spec: containers: - name: app resources: requests: cpu: "250m" limits: cpu: "500m"