Skip to content

fix: degrade dcgm query timeouts #222

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

Merged
jingxiang-z merged 1 commit into from
Jun 24, 2026
Merged

fix: degrade dcgm query timeouts #222

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
4 changes: 1 addition & 3 deletions third_party/fleet-intelligence-sdk/pkg/nvidia-query/dcgm/query_validator.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -210,9 +210,7 @@ func IsUnhealthyAPIError(err error) bool {
case dcgm.DCGM_ST_NVML_ERROR,
dcgm.DCGM_ST_GPU_IS_LOST,
dcgm.DCGM_ST_RESET_REQUIRED,
dcgm.DCGM_ST_GPU_NOT_SUPPORTED,
dcgm.DCGM_ST_TIMEOUT,
dcgm.DCGM_ST_NVML_DRIVER_TIMEOUT:
dcgm.DCGM_ST_GPU_NOT_SUPPORTED:
return true
default:
return false
Expand Down
28 changes: 28 additions & 0 deletions third_party/fleet-intelligence-sdk/pkg/nvidia-query/dcgm/query_validator_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,10 @@
package dcgm

import (
"fmt"
"testing"

dcgm "github.com/NVIDIA/go-dcgm/pkg/dcgm"
)

// Note: Tests for CheckSentinel() and CheckSentinelV2() require actual DCGM library
Expand Down Expand Up @@ -66,3 +69,28 @@ func TestSentinelType_String(t *testing.T) {
})
}
}

func TestIsUnhealthyAPIError(t *testing.T) {
tests := []struct {
name string
code int32
want bool
}{
{"nvml error is unhealthy", dcgm.DCGM_ST_NVML_ERROR, true},
{"gpu lost is unhealthy", dcgm.DCGM_ST_GPU_IS_LOST, true},
{"reset required is unhealthy", dcgm.DCGM_ST_RESET_REQUIRED, true},
{"gpu not supported is unhealthy", dcgm.DCGM_ST_GPU_NOT_SUPPORTED, true},
{"dcgm timeout is degraded", dcgm.DCGM_ST_TIMEOUT, false},
{"nvml driver timeout is degraded", dcgm.DCGM_ST_NVML_DRIVER_TIMEOUT, false},
{"stale data is not unhealthy", dcgm.DCGM_ST_STALE_DATA, false},
}

for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
err := fmt.Errorf("dcgm request failed with error code %d", tt.code)
if got := IsUnhealthyAPIError(err); got != tt.want {
t.Errorf("IsUnhealthyAPIError() = %v, want %v", got, tt.want)
}
})
}
}
Loading