reanimator/bee
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 1 Wiki Activity

Move NCCL and NVBandwidth into validate mode

Browse Source
This commit is contained in:
Mikhail Chusavitin
2026-04-16 11:02:30 +03:00
parent f74976ec4c
commit b4280941f5
7 changed files with 91 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
audit/internal/app/app.go
View File

@@ -146,7 +146,7 @@ type satRunner interface {
RunSATStressPack(ctx context.Context, baseDir string, durationSec int, logFunc func(string)) (string, error)
RunFanStressTest(ctx context.Context, baseDir string, opts platform.FanStressOptions) (string, error)
RunPlatformStress(ctx context.Context, baseDir string, opts platform.PlatformStressOptions, logFunc func(string)) (string, error)
RunNCCLTests(ctx context.Context, baseDir string, logFunc func(string)) (string, error)
RunNCCLTests(ctx context.Context, baseDir string, gpuIndices []int, logFunc func(string)) (string, error)
}
type runtimeChecker interface {
@@ -744,8 +744,15 @@ func (a *App) RunPlatformStress(ctx context.Context, baseDir string, opts platfo
return a.sat.RunPlatformStress(ctx, baseDir, opts, logFunc)
}
func (a *App) RunNCCLTests(ctx context.Context, baseDir string, gpuIndices []int, logFunc func(string)) (string, error) {
if strings.TrimSpace(baseDir) == "" {
baseDir = DefaultSATBaseDir
}
return a.sat.RunNCCLTests(ctx, baseDir, gpuIndices, logFunc)
}
func (a *App) RunNCCLTestsResult(ctx context.Context) (ActionResult, error) {
path, err := a.sat.RunNCCLTests(ctx, DefaultSATBaseDir, nil)
path, err := a.RunNCCLTests(ctx, DefaultSATBaseDir, nil, nil)
body := "Results: " + path
if err != nil && err != context.Canceled {
body += "\nERROR: " + err.Error()

36
audit/internal/app/app_test.go
View File

@@ -128,6 +128,7 @@ type fakeSAT struct {
runNvidiaPowerFn func(string, int, []int) (string, error)
runNvidiaPulseFn func(string, int, []int) (string, error)
runNvidiaBandwidthFn func(string, []int) (string, error)
runNCCLFn func(string, []int) (string, error)
runNvidiaTargetedStressFn func(string, int, []int) (string, error)
runMemoryFn func(string) (string, error)
runStorageFn func(string) (string, error)
@@ -287,10 +288,43 @@ func (f fakeSAT) RunPlatformStress(_ context.Context, _ string, _ platform.Platf
return "", nil
}
func (f fakeSAT) RunNCCLTests(_ context.Context, _ string, _ func(string)) (string, error) {
func (f fakeSAT) RunNCCLTests(_ context.Context, baseDir string, gpuIndices []int, _ func(string)) (string, error) {
if f.runNCCLFn != nil {
return f.runNCCLFn(baseDir, gpuIndices)
}
return "", nil
}
func TestRunNCCLTestsPassesSelectedGPUs(t *testing.T) {
t.Parallel()
var gotBaseDir string
var gotGPUIndices []int
a := &App{
sat: fakeSAT{
runNCCLFn: func(baseDir string, gpuIndices []int) (string, error) {
gotBaseDir = baseDir
gotGPUIndices = append([]int(nil), gpuIndices...)
return "/tmp/nccl-tests.tar.gz", nil
},
},
}
path, err := a.RunNCCLTests(context.Background(), "/tmp/sat", []int{3, 1}, nil)
if err != nil {
t.Fatalf("RunNCCLTests error: %v", err)
}
if path != "/tmp/nccl-tests.tar.gz" {
t.Fatalf("path=%q want %q", path, "/tmp/nccl-tests.tar.gz")
}
if gotBaseDir != "/tmp/sat" {
t.Fatalf("baseDir=%q want %q", gotBaseDir, "/tmp/sat")
}
if len(gotGPUIndices) != 2 || gotGPUIndices[0] != 3 || gotGPUIndices[1] != 1 {
t.Fatalf("gpuIndices=%v want [3 1]", gotGPUIndices)
}
}
func TestNetworkStatusFormatsInterfacesAndRoute(t *testing.T) {
t.Parallel()