Merge pull request #3774 from archlitchi/master

volcano-sh-bot · web-flow · commit 919df6908ddd · 2024-10-29T15:27:52.000+08:00
fix allocating more pods to a GPU when using volcano-vgpu feature
diff --git a/pkg/scheduler/api/devices/nvidia/vgpu/utils.go b/pkg/scheduler/api/devices/nvidia/vgpu/utils.go
@@ -359,7 +359,7 @@ func checkNodeGPUSharingPredicateAndScore(pod *v1.Pod, gssnap *GPUDevices, repli
 
 		for i := len(gs.Device) - 1; i >= 0; i-- {
 			klog.V(3).InfoS("Scoring pod request", "memReq", val.Memreq, "memPercentageReq", val.MemPercentagereq, "coresReq", val.Coresreq, "Nums", val.Nums, "Index", i, "ID", gs.Device[i].ID)
-			klog.V(3).InfoS("Current Device", "Index", i, "TotalMemory", gs.Device[i].Memory, "UsedMemory", gs.Device[i].UsedMem, "UsedCores", gs.Device[i].UsedNum)
+			klog.V(3).InfoS("Current Device", "Index", i, "TotalMemory", gs.Device[i].Memory, "UsedMemory", gs.Device[i].UsedMem, "UsedCores", gs.Device[i].UsedCore, "replicate", replicate)
 			if gs.Device[i].Number <= uint(gs.Device[i].UsedNum) {
 				continue
 			}
@@ -369,7 +369,7 @@ func checkNodeGPUSharingPredicateAndScore(pod *v1.Pod, gssnap *GPUDevices, repli
 			if int(gs.Device[i].Memory)-int(gs.Device[i].UsedMem) < int(val.Memreq) {
 				continue
 			}
-			if 100-gs.Device[i].UsedCore < uint(val.Coresreq) {
+			if 100-int32(gs.Device[i].UsedCore) < val.Coresreq {
 				continue
 			}
 			// Coresreq=100 indicates it want this card exclusively

Original file line number	Diff line number	Diff line change
`@@ -359,7 +359,7 @@ func checkNodeGPUSharingPredicateAndScore(pod v1.Pod, gssnap GPUDevices, repli`
`359`	`359`
`360`	`360`	`for i := len(gs.Device) - 1; i >= 0; i-- {`
`361`	`361`	`klog.V(3).InfoS("Scoring pod request", "memReq", val.Memreq, "memPercentageReq", val.MemPercentagereq, "coresReq", val.Coresreq, "Nums", val.Nums, "Index", i, "ID", gs.Device[i].ID)`
`362`		`- klog.V(3).InfoS("Current Device", "Index", i, "TotalMemory", gs.Device[i].Memory, "UsedMemory", gs.Device[i].UsedMem, "UsedCores", gs.Device[i].UsedNum)`
	`362`	`+ klog.V(3).InfoS("Current Device", "Index", i, "TotalMemory", gs.Device[i].Memory, "UsedMemory", gs.Device[i].UsedMem, "UsedCores", gs.Device[i].UsedCore, "replicate", replicate)`
`363`	`363`	`if gs.Device[i].Number <= uint(gs.Device[i].UsedNum) {`
`364`	`364`	`continue`
`365`	`365`	`}`
`@@ -369,7 +369,7 @@ func checkNodeGPUSharingPredicateAndScore(pod v1.Pod, gssnap GPUDevices, repli`
`369`	`369`	`if int(gs.Device[i].Memory)-int(gs.Device[i].UsedMem) < int(val.Memreq) {`
`370`	`370`	`continue`
`371`	`371`	`}`
`372`		`- if 100-gs.Device[i].UsedCore < uint(val.Coresreq) {`
	`372`	`+ if 100-int32(gs.Device[i].UsedCore) < val.Coresreq {`
`373`	`373`	`continue`
`374`	`374`	`}`
`375`	`375`	`// Coresreq=100 indicates it want this card exclusively`