Test works locally needed to add some metrics to dims

aws · Jun 10, 2024 · db41f67 · db41f67
1 parent 5855c50
commit db41f67
Show file tree

Hide file tree

Showing 3 changed files with 27 additions and 25 deletions.
diff --git a/test/gpu/nvidia_test.go b/test/gpu/nvidia_test.go
@@ -17,24 +17,28 @@ import (
 const (
 	gpuMetricIndicator = "_gpu_"
 
-	containerMemTotal   = "container_gpu_memory_total"
-	containerMemUsed    = "container_gpu_memory_used"
-	containerPower      = "container_gpu_power_draw"
-	containerTemp       = "container_gpu_temperature"
-	containerUtil       = "container_gpu_utilization"
-	containerMemUtil    = "container_gpu_memory_utilization"
-	podMemTotal         = "pod_gpu_memory_total"
-	podMemUsed          = "pod_gpu_memory_used"
-	podPower            = "pod_gpu_power_draw"
-	podTemp             = "pod_gpu_temperature"
-	podUtil             = "pod_gpu_utilization"
-	podMemUtil          = "pod_gpu_memory_utilization"
-	nodeMemTotal        = "node_gpu_memory_total"
-	nodeMemUsed         = "node_gpu_memory_used"
-	nodePower           = "node_gpu_power_draw"
-	nodeTemp            = "node_gpu_temperature"
-	nodeUtil            = "node_gpu_utilization"
-	nodeMemUtil         = "node_gpu_memory_utilization"
+	containerMemTotal = "container_gpu_memory_total"
+	containerMemUsed  = "container_gpu_memory_used"
+	containerPower    = "container_gpu_power_draw"
+	containerTemp     = "container_gpu_temperature"
+	containerUtil     = "container_gpu_utilization"
+	containerMemUtil  = "container_gpu_memory_utilization"
+	podMemTotal       = "pod_gpu_memory_total"
+	podMemUsed        = "pod_gpu_memory_used"
+	podPower          = "pod_gpu_power_draw"
+	podTemp           = "pod_gpu_temperature"
+	podUtil           = "pod_gpu_utilization"
+	podMemUtil        = "pod_gpu_memory_utilization"
+	podLimit          = "pod_gpu_limit"
+	podRequest        = "pod_gpu_request"
+	podTotal          = "pod_gpu_total"
+	nodeMemTotal      = "node_gpu_memory_total"
+	nodeMemUsed       = "node_gpu_memory_used"
+	nodePower         = "node_gpu_power_draw"
+	nodeTemp          = "node_gpu_temperature"
+	nodeUtil          = "node_gpu_utilization"
+	nodeMemUtil       = "node_gpu_memory_utilization"
+
 	nodeCountTotal      = "node_gpu_total"
 	nodeCountRequest    = "node_gpu_request"
 	nodeCountLimit      = "node_gpu_limit"
@@ -46,18 +50,16 @@ var expectedDimsToMetrics = map[string][]string{
 	"ClusterName": {
 		containerMemTotal, containerMemUsed, containerPower, containerTemp, containerUtil, containerMemUtil,
 		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil,
-		nodeMemTotal, nodeMemUsed, nodePower, nodeTemp, nodeUtil, nodeMemUtil,
-		//nodeCountTotal, nodeCountRequest, nodeCountLimit,
-		//clusterCountTotal, clusterCountRequest,
+		nodeMemTotal, nodeMemUsed, nodePower, nodeTemp, nodeUtil, nodeMemUtil, podLimit, podTotal, podRequest, nodeCountTotal, nodeCountRequest, nodeCountLimit, clusterCountTotal, clusterCountRequest,
 	},
 	"ClusterName-Namespace": {
-		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil,
+		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil, podLimit, podTotal, podRequest,
 	},
 	//"ClusterName-Namespace-Service": {
 	//	podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil,
 	//},
 	"ClusterName-Namespace-PodName": {
-		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil,
+		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil, podLimit, podTotal, podRequest,
 	},
 	"ClusterName-ContainerName-Namespace-PodName": {
 		containerMemTotal, containerMemUsed, containerPower, containerTemp, containerUtil, containerMemUtil,
@@ -69,7 +71,7 @@ var expectedDimsToMetrics = map[string][]string{
 		containerMemTotal, containerMemUsed, containerPower, containerTemp, containerUtil, containerMemUtil,
 	},
 	"ClusterName-FullPodName-Namespace-PodName": {
-		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil,
+		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil, podLimit, podTotal, podRequest,
 	},
 	"ClusterName-FullPodName-GpuDevice-Namespace-PodName": {
 		podMemTotal, podMemUsed, podPower, podTemp, podUtil, podMemUtil,

diff --git a/test/metric/container_insights_util.go b/test/metric/container_insights_util.go
@@ -57,6 +57,7 @@ func ValidateMetrics(env *environment.MetaData, metricFilter string, expectedDim
 			continue
 		}
 		results = append(results, validateMetricsAvailability(dims, metrics, actual))
+
 		for _, m := range metrics {
 			// this is to prevent panic with rand.Intn when metrics are not yet ready in a cluster
 			if _, ok := actual[m]; !ok {

diff --git a/test/metric_value_benchmark/eks_resources/test_schemas/pod_gpu.json b/test/metric_value_benchmark/eks_resources/test_schemas/pod_gpu.json
@@ -34,7 +34,6 @@
   "required": [
     "ClusterName",
     "FullPodName",
-    "GpuDevice",
     "InstanceId",
     "Namespace",
     "NodeName",