Skip to content

Expand NVIDIA device test coverage #2340

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
rishisulakhe wants to merge 2 commits into from
Closed

Expand NVIDIA device test coverage #2340

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a45a084
Expand NVIDIA device test coverage
rishisulakhe Aug 4, 2026
d38e065
fix gofmt
rishisulakhe Aug 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
206 changes: 205 additions & 1 deletion pkg/device/nvidia/device_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
/*/*
/*
Copyright 2024 The HAMi Authors.

Licensed under the Apache License, Version 2.0 (the "License");
Expand Down Expand Up @@ -2013,6 +2013,60 @@ func TestGenerateResourceRequests(t *testing.T) {
Coresreq: 0,
},
},
{
name: "gpu + mem in Requests only — falls back to Requests",
ctr: &corev1.Container{
Resources: corev1.ResourceRequirements{
Requests: corev1.ResourceList{
"nvidia.com/gpu": *resource.NewQuantity(1, resource.BinarySI),
"nvidia.com/gpumem": *resource.NewQuantity(2048, resource.DecimalSI),
},
},
},
want: device.ContainerDeviceRequest{
Nums: 1,
Type: NvidiaGPUDevice,
Memreq: 2048,
MemPercentagereq: 101,
Coresreq: 0,
},
},
{
name: "gpu + cores in Requests only — falls back to Requests",
ctr: &corev1.Container{
Resources: corev1.ResourceRequirements{
Requests: corev1.ResourceList{
"nvidia.com/gpu": *resource.NewQuantity(1, resource.BinarySI),
"nvidia.com/gpucores": *resource.NewQuantity(30, resource.DecimalSI),
},
},
},
want: device.ContainerDeviceRequest{
Nums: 1,
Type: NvidiaGPUDevice,
Memreq: 0,
MemPercentagereq: 100,
Coresreq: 30,
},
},
{
name: "gpu + mem-percentage in Requests only — falls back to Requests",
ctr: &corev1.Container{
Resources: corev1.ResourceRequirements{
Requests: corev1.ResourceList{
"nvidia.com/gpu": *resource.NewQuantity(2, resource.BinarySI),
"nvidia.com/gpumem-percentage": *resource.NewQuantity(25, resource.DecimalSI),
},
},
},
want: device.ContainerDeviceRequest{
Nums: 2,
Type: NvidiaGPUDevice,
Memreq: 0,
MemPercentagereq: 25,
Coresreq: 0,
},
},
}

for _, tt := range tests {
Expand Down Expand Up @@ -3042,3 +3096,153 @@ func TestFit_TopologyNegativeScores(t *testing.T) {
assert.Assert(t, uuids["dev-2"])
})
}

func TestMutateContainerResource(t *testing.T) {
tests := []struct {
name string
gpuNum int32
limits corev1.ResourceList
requests corev1.ResourceList
wantGPUAdded bool
expectedGPUNum int64
}{
{
name: "GPU count already in Limits — return true, no mutation",
gpuNum: 1,
limits: corev1.ResourceList{
"nvidia.com/gpu": *resource.NewQuantity(3, resource.BinarySI),
},
wantGPUAdded: true,
expectedGPUNum: 3,
},
{
name: "cores in Limits — injects default GPU count",
gpuNum: 1,
limits: corev1.ResourceList{
"nvidia.com/gpucores": *resource.NewQuantity(50, resource.DecimalSI),
},
wantGPUAdded: true,
expectedGPUNum: 1,
},
{
name: "memory in Limits — injects default GPU count",
gpuNum: 2,
limits: corev1.ResourceList{
"nvidia.com/gpumem": *resource.NewQuantity(4096, resource.DecimalSI),
},
wantGPUAdded: true,
expectedGPUNum: 2,
},
{
name: "mem percentage in Limits — injects default GPU count",
gpuNum: 1,
limits: corev1.ResourceList{
"nvidia.com/gpumem-percentage": *resource.NewQuantity(50, resource.DecimalSI),
},
wantGPUAdded: true,
expectedGPUNum: 1,
},
{
name: "DefaultGPUNum=0 with cores in Limits — no injection",
gpuNum: 0,
limits: corev1.ResourceList{
"nvidia.com/gpucores": *resource.NewQuantity(50, resource.DecimalSI),
},
wantGPUAdded: false,
},
{
name: "cores in Requests only — not seen (mutateContainerResource reads Limits only)",
gpuNum: 1,
requests: corev1.ResourceList{
"nvidia.com/gpucores": *resource.NewQuantity(50, resource.DecimalSI),
},
wantGPUAdded: false,
},
{
name: "GPU count in Requests only — not seen (mutateContainerResource reads Limits only)",
gpuNum: 1,
requests: corev1.ResourceList{
"nvidia.com/gpu": *resource.NewQuantity(4, resource.BinarySI),
},
wantGPUAdded: false,
},
{
name: "GPU count in Limits as zero — returns true",
gpuNum: 1,
limits: corev1.ResourceList{
"nvidia.com/gpu": *resource.NewQuantity(0, resource.BinarySI),
},
wantGPUAdded: true,
expectedGPUNum: 0,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
dev := &NvidiaGPUDevices{
config: NvidiaConfig{
ResourceCountName: "nvidia.com/gpu",
ResourceMemoryName: "nvidia.com/gpumem",
ResourceCoreName: "nvidia.com/gpucores",
ResourceMemoryPercentageName: "nvidia.com/gpumem-percentage",
DefaultGPUNum: tt.gpuNum,
},
}
ctr := &corev1.Container{
Resources: corev1.ResourceRequirements{
Limits: tt.limits,
Requests: tt.requests,
},
}
got := dev.mutateContainerResource(ctr)
assert.Equal(t, tt.wantGPUAdded, got)
if got {
gpuQty, ok := ctr.Resources.Limits["nvidia.com/gpu"]
assert.Assert(t, ok, "GPU limit should be present")
assert.Equal(t, tt.expectedGPUNum, gpuQty.Value())
}
})
}
}

func TestFit_MpsMode(t *testing.T) {
nv := InitNvidiaDevice(NvidiaConfig{
ResourceCountName: "nvidia.com/gpu",
ResourceMemoryName: "nvidia.com/gpumem",
ResourceCoreName: "nvidia.com/gpucores",
ResourceMemoryPercentageName: "nvidia.com/gpumem-percentage",
})
devices := []*device.DeviceUsage{
{ID: "dev-0", Index: 0, Used: 0, Count: 10, Totalmem: 8192, Totalcore: 100, Type: NvidiaGPUDevice, Health: true, Mode: "mps"},
{ID: "dev-1", Index: 1, Used: 0, Count: 10, Totalmem: 8192, Totalcore: 100, Type: NvidiaGPUDevice, Health: true, Mode: "mps"},
}

// MPS mode should allow allocation even when multiple containers share
req := device.ContainerDeviceRequest{Nums: 2, Memreq: 1000, Coresreq: 20, Type: NvidiaGPUDevice}
pod := &corev1.Pod{ObjectMeta: metav1.ObjectMeta{
Annotations: map[string]string{AllocateMode: "mps"},
}}
fit, result, _ := nv.Fit(devices, req, pod, &device.NodeInfo{}, &device.PodDevices{})
assert.Equal(t, true, fit)

@Eshiv-Pandey Eshiv-Pandey Aug 4, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

noticed a tiny inconsistency.. TestFit_NumaSwitching uses assert.Equal(t, fit, true), so consider sticking to tht argument order only. TestFit_MpsMode, TestFit_NumaBindDifferentNumaFails and TestMutateContainerResource make them follow the norm

assert.Equal(t, len(result[NvidiaGPUDevice]), 2)
}

func TestFit_NumaBindDifferentNumaFails(t *testing.T) {
nv := InitNvidiaDevice(NvidiaConfig{
ResourceCountName: "nvidia.com/gpu",
ResourceMemoryName: "nvidia.com/gpumem",
ResourceCoreName: "nvidia.com/gpucores",
ResourceMemoryPercentageName: "nvidia.com/gpumem-percentage",
})
devices := []*device.DeviceUsage{
{ID: "dev-0", Index: 0, Used: 0, Count: 10, Totalmem: 8192, Totalcore: 100, Type: NvidiaGPUDevice, Health: true, Numa: 0},
{ID: "dev-1", Index: 1, Used: 0, Count: 10, Totalmem: 8192, Totalcore: 100, Type: NvidiaGPUDevice, Health: true, Numa: 1},
}
req := device.ContainerDeviceRequest{Nums: 2, Memreq: 100, Coresreq: 10, Type: NvidiaGPUDevice}
pod := &corev1.Pod{ObjectMeta: metav1.ObjectMeta{
Annotations: map[string]string{NumaBind: "true"},
}}
fit, _, reason := nv.Fit(devices, req, pod, &device.NodeInfo{}, &device.PodDevices{})
assert.Equal(t, false, fit)
assert.Assert(t, strings.Contains(reason, "NumaNotFit"))
assert.Assert(t, strings.Contains(reason, "AllocatedCardsInsufficientRequest"))
}
Loading