AssignedNonTerminatedPod selects pods that are assigned and non-terminal (scheduled and running).

GetGpuAffinityFromPodAnnotation returns the gpu affinity of the pod.

GetGpuCountFromPodAnnotation gets the GPU Count of the pod.

GetGpuCountOfNode Get the GPU count of the node.

GetGpuIdFromAnnotation gets GPU ID from Annotation, could be "1" or "0-1-2-3" for multi-GPU allocations.

GetGpuIdListFromAnnotation gets GPU ID List from Annotation, could be [1] or [0, 1, 2, 3] for multi-GPU allocations.

GetGpuMilliFromPodAnnotation gets the GPU (in Milli) of the pod, range: 0-1000.

GetGpuMilliOfNode Get the GPU Milli of the node (0-8000 at most).

GetGpuModelOfNode Get the GPU Model of the node, see MapGpuTypeMemoryMiB.

GpuIdStrToIntList follows the string formed in func (n *GpuNodeInfo) AllocateGpuId.

IsCompletePod determines if the pod is complete.

IsGpuSharingNode Is the Node for GPU sharing.

UpdatePodDeviceAnnoSpec updates pod env with devId.

To retrieve the scheduling latency.

GPU number request (or allocatable), for pod and node.

creation timestamp.

deletion timestamp.

Exists when the pod are assigned/predefined to a GPU device.

CPU MILLI.

GPU MILLI.

Mem MiB.

GPU card model, for pod and node.

GPU milli, i.e., 1000 == 1 GPU, for pod only, node is 1000 by default.