Ci neuron

exporter/awsemfexporter/metric_translator.go

@@ -132,7 +132,8 @@ func (mt metricTranslator) translateOTelToGroupedMetric(rm pmetric.ResourceMetri
 
 	if serviceName, ok := rm.Resource().Attributes().Get("service.name"); ok {
 		if strings.HasPrefix(serviceName.Str(), "containerInsightsKubeAPIServerScraper") ||
-			strings.HasPrefix(serviceName.Str(), "containerInsightsDCGMExporterScraper") {
+			strings.HasPrefix(serviceName.Str(), "containerInsightsDCGMExporterScraper") ||
+			strings.HasPrefix(serviceName.Str(), "containerInsightsNeuronMonitorScraper") {
 			// the prometheus metrics that come from the container insight receiver need to be clearly tagged as coming from container insights
 			metricReceiver = containerInsightsReceiver
 		}

exporter/awsemfexporter/metric_translator_test.go

@@ -282,6 +282,8 @@ func TestTranslateOtToGroupedMetric(t *testing.T) {
 	containerInsightMetric.Resource().Attributes().PutStr(conventions.AttributeServiceName, "containerInsightsKubeAPIServerScraper")
 	gpuMetric := createTestResourceMetricsHelper(defaultNumberOfTestMetrics + 1)
 	gpuMetric.Resource().Attributes().PutStr(conventions.AttributeServiceName, "containerInsightsDCGMExporterScraper")
+	neuronMetric := createTestResourceMetricsHelper(defaultNumberOfTestMetrics + 1)
+	neuronMetric.Resource().Attributes().PutStr(conventions.AttributeServiceName, "containerInsightsNeuronMonitorScraper")
 
 	counterSumMetrics := map[string]*metricInfo{
 		"spanCounter": {
@@ -390,6 +392,20 @@ func TestTranslateOtToGroupedMetric(t *testing.T) {
 			"myServiceNS/containerInsightsDCGMExporterScraper",
 			containerInsightsReceiver,
 		},
+		{
+			"neuron monitor receiver",
+			neuronMetric,
+			map[string]string{
+				"isItAnError": "false",
+				"spanName":    "testSpan",
+			},
+			map[string]string{
+				oTellibDimensionKey: "cloudwatch-lib",
+				"spanName":          "testSpan",
+			},
+			"myServiceNS/containerInsightsNeuronMonitorScraper",
+			containerInsightsReceiver,
+		},
 	}
 
 	for _, tc := range testCases {

internal/aws/containerinsight/const.go

@@ -150,11 +150,12 @@ const (
 	TypeContainerDiskIO    = "ContainerDiskIO"
 	// Special type for pause container
 	// because containerd does not set container name pause container name to POD like docker does.
-	TypeInfraContainer = "InfraContainer"
-	TypeGpuContainer   = "ContainerGPU"
-	TypeGpuPod         = "PodGPU"
-	TypeGpuNode        = "NodeGPU"
-	TypeGpuCluster     = "ClusterGPU"
+	TypeInfraContainer  = "InfraContainer"
+	TypeGpuContainer    = "ContainerGPU"
+	TypeGpuPod          = "PodGPU"
+	TypeGpuNode         = "NodeGPU"
+	TypeGpuCluster      = "ClusterGPU"
+	TypeNeuronContainer = "ContainerNeuron"
 
 	// unit
 	UnitBytes       = "Bytes"

internal/aws/proxy/server_test.go

@@ -226,7 +226,7 @@ func TestCanCreateTransport(t *testing.T) {
 
 	_, err := NewServer(cfg, logger)
 	assert.Error(t, err, "NewServer should fail")
-	assert.Contains(t, err.Error(), "failed to parse proxy URL")
+	assert.Contains(t, err.Error(), "invalid control character in URL")
 }
 
 func TestGetServiceEndpointInvalidAWSConfig(t *testing.T) {

receiver/awscontainerinsightreceiver/config.go

@@ -60,4 +60,9 @@ type Config struct {
 
 	// EnableAcceleratedComputeMetrics enabled features with accelerated compute resources where metrics are scraped from vendor specific sources
 	EnableAcceleratedComputeMetrics bool `mapstructure:"accelerated_compute_metrics"`
+	EnableGpuMetric                 bool `mapstructure:"accelerated_compute_metrics"`
+
+	// EnableNeuronMetric toggles Neuron monitoring where metrics are scraped from neuron monitor
+	// The default value is false.
+	EnableNeuronMetric bool `mapstructure:"neuron_metrics"`
 }

receiver/awscontainerinsightreceiver/factory.go

@@ -42,6 +42,9 @@ const (
 
 	// Don't enable EKS control plane metrics by default
 	defaultEnableControlPlaneMetrics = false
+
+	// Don't enable Neuron metrics by default
+	defaultEnableNeuronMetrics = false
 )
 
 // NewFactory creates a factory for AWS container insight receiver
@@ -64,6 +67,7 @@ func createDefaultConfig() component.Config {
 		ClusterName:                 defaultClusterName,
 		LeaderLockName:              defaultLeaderLockName,
 		EnableControlPlaneMetrics:   defaultEnableControlPlaneMetrics,
+		EnableNeuronMetric:          defaultEnableNeuronMetrics,
 	}
 }
 

receiver/awscontainerinsightreceiver/go.mod

@@ -26,11 +26,13 @@ require (
 	go.opentelemetry.io/collector/receiver v0.89.0
 	go.uber.org/zap v1.26.0
 	golang.org/x/exp v0.0.0-20231127185646-65229373498e
+	google.golang.org/grpc v1.59.0
 	gopkg.in/yaml.v2 v2.4.0
 	k8s.io/api v0.28.3
 	k8s.io/apimachinery v0.28.3
 	k8s.io/client-go v0.28.3
 	k8s.io/klog v1.0.0
+	k8s.io/kubelet v0.27.3
 	k8s.io/utils v0.0.0-20230711102312-30195339c3c7
 )
 
@@ -210,7 +212,6 @@ require (
 	google.golang.org/genproto v0.0.0-20231030173426-d783a09b4405 // indirect
 	google.golang.org/genproto/googleapis/api v0.0.0-20231106174013-bbf56f31fb17 // indirect
 	google.golang.org/genproto/googleapis/rpc v0.0.0-20231030173426-d783a09b4405 // indirect
-	google.golang.org/grpc v1.59.0 // indirect
 	google.golang.org/protobuf v1.31.0 // indirect
 	gopkg.in/inf.v0 v0.9.1 // indirect
 	gopkg.in/ini.v1 v1.67.0 // indirect

receiver/awscontainerinsightreceiver/internal/gpu/dcgmscraper.go

@@ -9,6 +9,7 @@ import (
 	"fmt"
 	"time"
 
+	configutil "github.com/prometheus/common/config"
 	"github.com/prometheus/common/model"
 	"github.com/prometheus/prometheus/config"
 	"github.com/prometheus/prometheus/discovery"
@@ -25,7 +26,7 @@ import (
 )
 
 const (
-	caFile                    = "/var/run/secrets/kubernetes.io/serviceaccount/ca.crt"
+	caFile                    = "/etc/amazon-cloudwatch-observability-agent-cert/tls-ca.crt"
 	collectionInterval        = 60 * time.Second
 	jobName                   = "containerInsightsDCGMExporterScraper"
 	scraperMetricsPath        = "/metrics"
@@ -103,10 +104,16 @@ func NewDcgmScraper(opts DcgmScraperOpts) (*DcgmScraper, error) {
 
 func getScraperConfig(hostInfoProvider hostInfoProvider) *config.ScrapeConfig {
 	return &config.ScrapeConfig{
+		HTTPClientConfig: configutil.HTTPClientConfig{
+			TLSConfig: configutil.TLSConfig{
+				CAFile:             caFile,
+				InsecureSkipVerify: false,
+			},
+		},
 		ScrapeInterval: model.Duration(collectionInterval),
 		ScrapeTimeout:  model.Duration(collectionInterval),
 		JobName:        jobName,
-		Scheme:         "http",
+		Scheme:         "https",
 		MetricsPath:    scraperMetricsPath,
 		ServiceDiscoveryConfigs: discovery.Configs{
 			&kubernetes.SDConfig{

receiver/awscontainerinsightreceiver/internal/gpu/dcgmscraper_config.go

@@ -0,0 +1,149 @@
+// Copyright The OpenTelemetry Authors
+// SPDX-License-Identifier: Apache-2.0
+
+package gpu
+
+import (
+	"context"
+	"time"
+
+	"github.com/prometheus/common/model"
+	"github.com/prometheus/prometheus/config"
+	"github.com/prometheus/prometheus/discovery"
+	"github.com/prometheus/prometheus/discovery/kubernetes"
+	"github.com/prometheus/prometheus/model/relabel"
+	"go.opentelemetry.io/collector/component"
+	"go.opentelemetry.io/collector/consumer"
+	"go.opentelemetry.io/collector/receiver"
+	"go.uber.org/zap"
+
+	ci "github.com/open-telemetry/opentelemetry-collector-contrib/internal/aws/containerinsight"
+	"github.com/open-telemetry/opentelemetry-collector-contrib/receiver/awscontainerinsightreceiver/internal/prometheusscraper"
+)
+
+const (
+	caFile                    = "/var/run/secrets/kubernetes.io/serviceaccount/ca.crt"
+	collectionInterval        = 60 * time.Second
+	jobName                   = "containerInsightsDCGMExporterScraper"
+	scraperMetricsPath        = "/metrics"
+	scraperK8sServiceSelector = "k8s-app=dcgm-exporter-service"
+)
+
+type DcgmScraper struct {
+	ctx                context.Context
+	settings           component.TelemetrySettings
+	host               component.Host
+	hostInfoProvider   hostInfoProvider
+	prometheusReceiver receiver.Metrics
+	k8sDecorator       prometheusscraper.Decorator
+	running            bool
+}
+
+type DcgmScraperOpts struct {
+	Ctx               context.Context
+	TelemetrySettings component.TelemetrySettings
+	Consumer          consumer.Metrics
+	Host              component.Host
+	HostInfoProvider  hostInfoProvider
+	K8sDecorator      prometheusscraper.Decorator
+	Logger            *zap.Logger
+}
+
+type hostInfoProvider interface {
+	GetClusterName() string
+	GetInstanceID() string
+}
+
+func GetScraperConfig(hostInfoProvider hostInfoProvider) *config.ScrapeConfig {
+	return &config.ScrapeConfig{
+		ScrapeInterval: model.Duration(collectionInterval),
+		ScrapeTimeout:  model.Duration(collectionInterval),
+		JobName:        jobName,
+		Scheme:         "http",
+		MetricsPath:    scraperMetricsPath,
+		ServiceDiscoveryConfigs: discovery.Configs{
+			&kubernetes.SDConfig{
+				Role: kubernetes.RoleService,
+				NamespaceDiscovery: kubernetes.NamespaceDiscovery{
+					IncludeOwnNamespace: true,
+				},
+				Selectors: []kubernetes.SelectorConfig{
+					{
+						Role:  kubernetes.RoleService,
+						Label: scraperK8sServiceSelector,
+					},
+				},
+			},
+		},
+		MetricRelabelConfigs: getMetricRelabelConfig(hostInfoProvider),
+	}
+}
+
+func getMetricRelabelConfig(hostInfoProvider hostInfoProvider) []*relabel.Config {
+	return []*relabel.Config{
+		{
+			SourceLabels: model.LabelNames{"__name__"},
+			Regex:        relabel.MustNewRegexp("DCGM_.*"),
+			Action:       relabel.Keep,
+		},
+		{
+			SourceLabels: model.LabelNames{"Hostname"},
+			TargetLabel:  ci.NodeNameKey,
+			Regex:        relabel.MustNewRegexp("(.*)"),
+			Replacement:  "${1}",
+			Action:       relabel.Replace,
+		},
+		{
+			SourceLabels: model.LabelNames{"namespace"},
+			TargetLabel:  ci.AttributeK8sNamespace,
+			Regex:        relabel.MustNewRegexp("(.*)"),
+			Replacement:  "${1}",
+			Action:       relabel.Replace,
+		},
+		// hacky way to inject static values (clusterName & instanceId) to label set without additional processor
+		// relabel looks up an existing label then creates another label with given key (TargetLabel) and value (static)
+		{
+			SourceLabels: model.LabelNames{"namespace"},
+			TargetLabel:  ci.ClusterNameKey,
+			Regex:        relabel.MustNewRegexp(".*"),
+			Replacement:  hostInfoProvider.GetClusterName(),
+			Action:       relabel.Replace,
+		},
+		{
+			SourceLabels: model.LabelNames{"namespace"},
+			TargetLabel:  ci.InstanceID,
+			Regex:        relabel.MustNewRegexp(".*"),
+			Replacement:  hostInfoProvider.GetInstanceID(),
+			Action:       relabel.Replace,
+		},
+		{
+			SourceLabels: model.LabelNames{"pod"},
+			TargetLabel:  ci.AttributeFullPodName,
+			Regex:        relabel.MustNewRegexp("(.*)"),
+			Replacement:  "${1}",
+			Action:       relabel.Replace,
+		},
+		// additional k8s podname for service name and k8s blob decoration
+		{
+			SourceLabels: model.LabelNames{"pod"},
+			TargetLabel:  ci.AttributeK8sPodName,
+			Regex:        relabel.MustNewRegexp("(.*)"),
+			Replacement:  "${1}",
+			Action:       relabel.Replace,
+		},
+		{
+			SourceLabels: model.LabelNames{"container"},
+			TargetLabel:  ci.AttributeContainerName,
+			Regex:        relabel.MustNewRegexp("(.*)"),
+			Replacement:  "${1}",
+			Action:       relabel.Replace,
+		},
+		{
+			SourceLabels: model.LabelNames{"device"},
+			TargetLabel:  ci.AttributeGpuDevice,
+			Regex:        relabel.MustNewRegexp("(.*)"),
+			Replacement:  "${1}",
+			Action:       relabel.Replace,
+		},
+	}
+}