supprot mgpu feature for volcano

go.mod

@@ -1,6 +1,6 @@
 module volcano.sh/volcano
 
-go 1.20
+go 1.19
 
 require (
 	github.com/agiledragon/gomonkey/v2 v2.1.0

pkg/scheduler/api/devices/nvidia/gpushare/device_info.go

@@ -91,6 +91,10 @@ func NewGPUDevices(name string, node *v1.Node) *GPUDevices {
 	return &gpudevices
 }
 
+func (gs *GPUDevices) DeepCopy() interface{} {
+	return gs
+}
+
 // GetIgnoredDevices return device names which wish vc-scheduler to ignore
 func (gs *GPUDevices) GetIgnoredDevices() []string {
 	return []string{""}
@@ -148,7 +152,7 @@ func (gs *GPUDevices) Release(kubeClient kubernetes.Interface, pod *v1.Pod) erro
 	return nil
 }
 
-func (gs *GPUDevices) FilterNode(pod *v1.Pod) (int, string, error) {
+func (gs *GPUDevices) FilterNode(kubeClient kubernetes.Interface, pod *v1.Pod) (int, string, error) {
 	klog.V(4).Infoln("DeviceSharing:Into FitInPod", pod.Name)
 	if GpuSharingEnable {
 		fit, err := checkNodeGPUSharingPredicate(pod, gs)

pkg/scheduler/api/devices/nvidia/mgpu/constants.go

@@ -0,0 +1,110 @@
+package mgpu
+
+import v1 "k8s.io/api/core/v1"
+
+/*
+	   actions: "reclaim, allocate, backfill, preempt"
+	   tiers:
+	   - plugins:
+	     - name: priority
+	     - name: gang
+	     - name: conformance
+	   - plugins:
+	     - name: drf
+	     - name: predicates
+	       arguments:
+	         predicate.MGPUEnable: true
+		     predicate.MGPUPolicy: binpack
+	 		 predicate.MGPUWeightOfCore: 20
+			 predicate.MGPUScheduleMode: index
+			 predicate.MGPUMaxContainersPerCard: 16
+	     - name: proportion
+	     - name: nodeorder
+*/
+
+var (
+	MGPUEnable   bool
+	GlobalConfig = &MGPUConfig{
+		Policy:                      Binpack,
+		MaxContainersNumPolicyValid: DefaultMaxContainersNumPolicyValid,
+		WeightOfCore:                DefaultWeightOfCore,
+		ScheduleMode:                DefaultScheduleMode,
+		MaxContainersPerCard:        DefaultMaxContainersPerCard,
+	}
+)
+
+type MGPUConfig struct {
+	// policy is the card-level scoring policy, e.g. binpack/spread
+	Policy string
+	// MaxContainersNumPolicyValid means that bigger then the number GPU Policy is invalid
+	MaxContainersNumPolicyValid int
+	// WeightOfCore is the weight of core, e.g. 20
+	WeightOfCore int
+	// ScheduleMode is the schedule mode, e.g. index/id
+	ScheduleMode string
+	// MaxContainersPerCard is the number of max containers per card, e.g. 16
+	MaxContainersPerCard int
+}
+
+const (
+	Policy                             = "predicate.MGPUPolicy"
+	MaxContainersNumPolicyValid        = "predicate.MGPUMaxContainersNumPolicyValid"
+	WeightOfCore                       = "predicate.MGPUWeightOfCore"
+	ScheduleMode                       = "predicate.MGPUScheduleMode"
+	MaxContainersPerCard               = "predicate.MGPUMaxContainersPerCard"
+	DefaultWeightOfCore                = 20
+	DefaultMaxContainersNumPolicyValid = 5
+	DefaultMaxContainersPerCard        = 16
+	DefaultScheduleMode                = "index"
+)
+
+// ModeType is a type "string".
+type ModeType string
+
+// PolicyType is a type "string".
+type PolicyType string
+
+const (
+	// Binpack is the string "binpack".
+	Binpack = "binpack"
+	// Spread is the string "spread".
+	Spread = "spread"
+)
+
+const (
+	// GPUCoreEachCard is GPUCoreEachCard
+	GPUCoreEachCard = 100
+	// NotNeedGPU is NotNeedGPU
+	NotNeedGPU = -1
+	// NotNeedRate is NotNeedRate
+	NotNeedRate = -2
+	// NotNeedMultipleGPU is NotNeedMultipleGPU
+	NotNeedMultipleGPU = -3
+	// GPUTypeMGPU is GPUTypeMGPU
+	GPUTypeMGPU = "mgpu"
+	// GPUTypeNvidiaGPU is GPUTypeNvidiaGPU
+	GPUTypeNvidiaGPU = "nvidia"
+	// DefaultComputePolicy is DefaultComputePolicy
+	DefaultComputePolicy = "fixed-share"
+	// NativeBurstSharePolicy is NativeBurstSharePolicy
+	NativeBurstSharePolicy = "native-burst-share"
+	// DefaultGPUCount is DefaultGPUCount
+	DefaultGPUCount = 1
+
+	// VKEAnnotationMGPUAssumed is the scheduled result for mgpu pod
+	VKEAnnotationMGPUAssumed = "vke.volcengine.com/assumed"
+	// VKEAnnotationMGPUContainer is the format for gpu index with specific container name
+	VKEAnnotationMGPUContainer = "vke.volcengine.com/gpu-index-container-%s"
+	// VKELabelNodeResourceType is label key of node resource type, the value is like nvidia/mgpu
+	VKELabelNodeResourceType = "vke.node.gpu.schedule"
+	// VKEAnnotationMGPUComputePolicy is annotation for mgpu compute policy on vke
+	VKEAnnotationMGPUComputePolicy = "vke.volcengine.com/mgpu-compute-policy"
+	// VKEAnnotationContainerMultipleGPU is annotation for mgpu container multiple mode on vke
+	VKEAnnotationContainerMultipleGPU = "vke.volcengine.com/container-multiple-gpu"
+	// VKEResourceMGPUCore is mgpu-core resource on vke
+	VKEResourceMGPUCore v1.ResourceName = "vke.volcengine.com/mgpu-core"
+	// VKEResourceMGPUMemory is mgpu-memory resource on vke
+	VKEResourceMGPUMemory v1.ResourceName = "vke.volcengine.com/mgpu-memory"
+
+	DeviceName = "mgpu"
+)