Add support for MIG and vGPUs in exporter

Makefile

@@ -151,6 +151,7 @@ test-e2e: build pkg/collector/testdata/sys/.unpacked pkg/collector/testdata/proc
 	./scripts/e2e-test.sh -s exporter-cgroups-v1
 	./scripts/e2e-test.sh -s exporter-cgroups-v1-memory-subsystem
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-nvidia-ipmiutil
+	./scripts/e2e-test.sh -s exporter-cgroups-v2-nvidia-gpu-reordering
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-amd-ipmitool
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-nogpu
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-procfs
@@ -202,6 +203,7 @@ test-e2e-update: build pkg/collector/testdata/sys/.unpacked pkg/collector/testda
 	./scripts/e2e-test.sh -s exporter-cgroups-v1 -u || true
 	./scripts/e2e-test.sh -s exporter-cgroups-v1-memory-subsystem -u || true
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-nvidia-ipmiutil -u || true
+	./scripts/e2e-test.sh -s exporter-cgroups-v2-nvidia-gpu-reordering -u || true
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-amd-ipmitool -u || true
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-nogpu -u || true
 	./scripts/e2e-test.sh -s exporter-cgroups-v2-procfs -u || true

README.md

@@ -35,7 +35,7 @@ in a resource manager agnostic way.
 
 - Monitor energy, performance, IO and network metrics for different types of resource
 managers (SLURM, Openstack, k8s)
-- Support NVIDIA and AMD GPUs
+- Support NVIDIA (MIG and vGPU) and AMD GPUs
 - Realtime access to metrics *via* Grafana dashboards
 - Access control to Prometheus datasource in Grafana
 - Stores aggregated metrics in a separate DB that can be retained for long time

cmd/ceems_api_server/main_test.go

@@ -7,6 +7,8 @@ import (
 	"path/filepath"
 	"testing"
 	"time"
+
+	"github.com/stretchr/testify/require"
 )
 
 var binary, _ = filepath.Abs("../../bin/ceems_api_server")
@@ -24,23 +26,17 @@ func TestBatchjobStatsExecutable(t *testing.T) {
 	tmpSacctPath := tmpDir + "/sacct"
 
 	sacctPath, err := filepath.Abs("../../pkg/api/testdata/sacct")
-	if err != nil {
-		t.Error(err)
-	}
+	require.NoError(t, err)
 
 	err = os.Link(sacctPath, tmpSacctPath)
-	if err != nil {
-		t.Error(err)
-	}
+	require.NoError(t, err)
 
 	usagestats := exec.Command(
 		binary,
 		"--web.listen-address", address,
 		"--no-security.drop-privileges",
 	)
-	if err := runCommandAndTests(usagestats); err != nil {
-		t.Error(err)
-	}
+	require.NoError(t, runCommandAndTests(usagestats))
 }
 
 func runCommandAndTests(cmd *exec.Cmd) error {

cmd/ceems_exporter/main_test.go

@@ -11,6 +11,7 @@ import (
 	"time"
 
 	"github.com/prometheus/procfs"
+	"github.com/stretchr/testify/require"
 )
 
 var binary, _ = filepath.Abs("../../bin/ceems_exporter")
@@ -37,14 +38,10 @@ func TestFileDescriptorLeak(t *testing.T) {
 	}
 
 	sysfsPath, err := filepath.Abs("../../pkg/collector/testdata/sys/fs/cgroup")
-	if err != nil {
-		t.Errorf("Failed to read testdata: %s", err)
-	}
+	require.NoError(t, err)
 
 	procfsPath, err := filepath.Abs("../../pkg/collector/testdata/proc")
-	if err != nil {
-		t.Errorf("Failed to read testdata: %s", err)
-	}
+	require.NoError(t, err)
 
 	exporter := exec.Command(
 		binary,
@@ -91,9 +88,7 @@ func TestFileDescriptorLeak(t *testing.T) {
 		return nil
 	}
 
-	if err := runCommandAndTests(exporter, address, test); err != nil {
-		t.Error(err)
-	}
+	require.NoError(t, runCommandAndTests(exporter, address, test))
 }
 
 func queryExporter(address string) error {

cmd/ceems_lb/main_test.go

@@ -7,6 +7,8 @@ import (
 	"path/filepath"
 	"testing"
 	"time"
+
+	"github.com/stretchr/testify/require"
 )
 
 var binary, _ = filepath.Abs("../../bin/ceems_lb")
@@ -24,24 +26,18 @@ func TestCEEMSLBExecutable(t *testing.T) {
 	tmpConfigPath := tmpDir + "/config.yaml"
 
 	configPath, err := filepath.Abs("../../build/config/ceems_lb/ceems_lb.yml")
-	if err != nil {
-		t.Error(err)
-	}
+	require.NoError(t, err)
 
 	err = os.Link(configPath, tmpConfigPath)
-	if err != nil {
-		t.Error(err)
-	}
+	require.NoError(t, err)
 
 	lb := exec.Command(
 		binary, "--path.data", tmpDir,
 		"--config.path", tmpConfigPath,
 		"--web.listen-address", address,
 		"--no-security.drop-privileges",
 	)
-	if err := runCommandAndTests(lb); err != nil {
-		t.Error(err)
-	}
+	require.NoError(t, runCommandAndTests(lb))
 }
 
 func runCommandAndTests(cmd *exec.Cmd) error {