discovery: enable by default on agents >= 7.78

api/datadoghq/v2alpha1/datadogagent_types.go

@@ -643,7 +643,8 @@ type USMFeatureConfig struct {
 // ServiceDiscoveryFeatureConfig configures the service discovery check feature.
 type ServiceDiscoveryFeatureConfig struct {
 	// Enables the service discovery check.
-	// Default: false
+	// Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false.
+	// If the image version cannot be determined, it is treated as latest.
 	// +optional
 	Enabled *bool `json:"enabled,omitempty"`
 

config/crd/bases/v1/datadoghq.com_datadogagentinternals.yaml

@@ -2501,7 +2501,8 @@ spec:
                         enabled:
                           description: |-
                             Enables the service discovery check.
-                            Default: false
+                            Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false.
+                            If the image version cannot be determined, it is treated as latest.
                           type: boolean
                         networkStats:
                           description: 'DEPRECATED: NetworkStats is no longer configurable and will be ignored. Scheduled for removal in v1.28.'
@@ -10933,7 +10934,8 @@ spec:
                             enabled:
                               description: |-
                                 Enables the service discovery check.
-                                Default: false
+                                Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false.
+                                If the image version cannot be determined, it is treated as latest.
                               type: boolean
                             networkStats:
                               description: 'DEPRECATED: NetworkStats is no longer configurable and will be ignored. Scheduled for removal in v1.28.'

config/crd/bases/v1/datadoghq.com_datadogagentinternals_v1alpha1.json

@@ -2600,7 +2600,7 @@
               "description": "ServiceDiscovery",
               "properties": {
                 "enabled": {
-                  "description": "Enables the service discovery check.\nDefault: false",
+                  "description": "Enables the service discovery check.\nDefault: true when omitted and the node Agent image is \u003e= 7.78.0. Otherwise false.\nIf the image version cannot be determined, it is treated as latest.",
                   "type": "boolean"
                 },
                 "networkStats": {
@@ -10769,7 +10769,7 @@
                   "description": "ServiceDiscovery",
                   "properties": {
                     "enabled": {
-                      "description": "Enables the service discovery check.\nDefault: false",
+                      "description": "Enables the service discovery check.\nDefault: true when omitted and the node Agent image is \u003e= 7.78.0. Otherwise false.\nIf the image version cannot be determined, it is treated as latest.",
                       "type": "boolean"
                     },
                     "networkStats": {

config/crd/bases/v1/datadoghq.com_datadogagentprofiles.yaml

@@ -2501,7 +2501,8 @@ spec:
                             enabled:
                               description: |-
                                 Enables the service discovery check.
-                                Default: false
+                                Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false.
+                                If the image version cannot be determined, it is treated as latest.
                               type: boolean
                             networkStats:
                               description: 'DEPRECATED: NetworkStats is no longer configurable and will be ignored. Scheduled for removal in v1.28.'

config/crd/bases/v1/datadoghq.com_datadogagentprofiles_v1alpha1.json

@@ -2604,7 +2604,7 @@
                   "description": "ServiceDiscovery",
                   "properties": {
                     "enabled": {
-                      "description": "Enables the service discovery check.\nDefault: false",
+                      "description": "Enables the service discovery check.\nDefault: true when omitted and the node Agent image is \u003e= 7.78.0. Otherwise false.\nIf the image version cannot be determined, it is treated as latest.",
                       "type": "boolean"
                     },
                     "networkStats": {

config/crd/bases/v1/datadoghq.com_datadogagents.yaml

@@ -2505,7 +2505,8 @@ spec:
                         enabled:
                           description: |-
                             Enables the service discovery check.
-                            Default: false
+                            Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false.
+                            If the image version cannot be determined, it is treated as latest.
                           type: boolean
                         networkStats:
                           description: 'DEPRECATED: NetworkStats is no longer configurable and will be ignored. Scheduled for removal in v1.28.'
@@ -11013,7 +11014,8 @@ spec:
                             enabled:
                               description: |-
                                 Enables the service discovery check.
-                                Default: false
+                                Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false.
+                                If the image version cannot be determined, it is treated as latest.
                               type: boolean
                             networkStats:
                               description: 'DEPRECATED: NetworkStats is no longer configurable and will be ignored. Scheduled for removal in v1.28.'

config/crd/bases/v1/datadoghq.com_datadogagents_v2alpha1.json

@@ -2600,7 +2600,7 @@
               "description": "ServiceDiscovery",
               "properties": {
                 "enabled": {
-                  "description": "Enables the service discovery check.\nDefault: false",
+                  "description": "Enables the service discovery check.\nDefault: true when omitted and the node Agent image is \u003e= 7.78.0. Otherwise false.\nIf the image version cannot be determined, it is treated as latest.",
                   "type": "boolean"
                 },
                 "networkStats": {
@@ -10862,7 +10862,7 @@
                   "description": "ServiceDiscovery",
                   "properties": {
                     "enabled": {
-                      "description": "Enables the service discovery check.\nDefault: false",
+                      "description": "Enables the service discovery check.\nDefault: true when omitted and the node Agent image is \u003e= 7.78.0. Otherwise false.\nIf the image version cannot be determined, it is treated as latest.",
                       "type": "boolean"
                     },
                     "networkStats": {

docs/configuration.v2alpha1.md

@@ -201,7 +201,7 @@ spec:
 | features.sbom.enabled | Enable this option to activate SBOM collection. Default: false |
 | features.sbom.host.analyzers | To use for SBOM collection. |
 | features.sbom.host.enabled | Enable this option to activate SBOM collection. Default: false |
-| features.serviceDiscovery.enabled | Enables the service discovery check. Default: false |
+| features.serviceDiscovery.enabled | Enables the service discovery check. Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false. If the image version cannot be determined, it is treated as latest. |
 | features.serviceDiscovery.networkStats.enabled | DEPRECATED: this field is ignored. |
 | features.tcpQueueLength.enabled | Enables the TCP queue length eBPF-based check. Default: false |
 | features.usm.enabled | Enables Universal Service Monitoring. Default: false |

docs/configuration_public.md

@@ -382,7 +382,7 @@ spec:
 : Enable this option to activate SBOM collection. Default: false
 
 `features.serviceDiscovery.enabled`
-: Enables the service discovery check. Default: false
+: Enables the service discovery check. Default: true when omitted and the node Agent image is >= 7.78.0. Otherwise false. If the image version cannot be determined, it is treated as latest.
 
 `features.serviceDiscovery.networkStats.enabled`
 : DEPRECATED: this field is ignored.

internal/controller/datadogagent/controller_v2_test.go

@@ -411,6 +411,106 @@ func TestReconcileDatadogAgentV2_Reconcile(t *testing.T) {
 				verifyDaemonsetContainers(t, c, resourcesNamespace, dsName, expectedContainers)
 			},
 		},
+		{
+			name: "DatadogAgent with service discovery omitted and supported node agent version auto-enables system-probe",
+			loadFunc: func(c client.Client) *v2alpha1.DatadogAgent {
+				dda := testutils.NewInitializedDatadogAgentBuilder(resourcesNamespace, resourcesName).
+					WithComponentOverride(v2alpha1.NodeAgentComponentName, v2alpha1.DatadogAgentComponentOverride{
+						Image: &v2alpha1.AgentImageConfig{Tag: "7.78.0"},
+					}).
+					Build()
+				_ = c.Create(context.TODO(), dda)
+				return dda
+			},
+			want:    reconcile.Result{RequeueAfter: defaultRequeueDuration},
+			wantErr: false,
+			wantFunc: func(t *testing.T, c client.Client) {
+				expectedContainers := []string{
+					string(apicommon.CoreAgentContainerName),
+					string(apicommon.TraceAgentContainerName),
+					string(apicommon.SystemProbeContainerName),
+				}
+
+				verifyDaemonsetContainers(t, c, resourcesNamespace, dsName, expectedContainers)
+
+				agentContainers := getDsContainers(c, resourcesNamespace, dsName)
+				assertContainerHasEnvVar(t, agentContainers[apicommon.CoreAgentContainerName], "DD_DISCOVERY_ENABLED", "true")
+				assertContainerHasEnvVar(t, agentContainers[apicommon.SystemProbeContainerName], common.DDSystemProbeSocket, common.DefaultSystemProbeSocketPath)
+			},
+		},
+		{
+			name: "DatadogAgent with service discovery omitted and unsupported node agent version keeps system-probe disabled",
+			loadFunc: func(c client.Client) *v2alpha1.DatadogAgent {
+				dda := testutils.NewInitializedDatadogAgentBuilder(resourcesNamespace, resourcesName).
+					WithComponentOverride(v2alpha1.NodeAgentComponentName, v2alpha1.DatadogAgentComponentOverride{
+						Image: &v2alpha1.AgentImageConfig{Tag: "7.77.2"},
+					}).
+					Build()
+				_ = c.Create(context.TODO(), dda)
+				return dda
+			},
+			want:    reconcile.Result{RequeueAfter: defaultRequeueDuration},
+			wantErr: false,
+			wantFunc: func(t *testing.T, c client.Client) {
+				expectedContainers := []string{
+					string(apicommon.CoreAgentContainerName),
+					string(apicommon.TraceAgentContainerName),
+				}
+
+				verifyDaemonsetContainers(t, c, resourcesNamespace, dsName, expectedContainers)
+			},
+		},
+		{
+			name: "DatadogAgent with service discovery explicitly disabled keeps system-probe disabled",
+			loadFunc: func(c client.Client) *v2alpha1.DatadogAgent {
+				dda := testutils.NewInitializedDatadogAgentBuilder(resourcesNamespace, resourcesName).
+					WithComponentOverride(v2alpha1.NodeAgentComponentName, v2alpha1.DatadogAgentComponentOverride{
+						Image: &v2alpha1.AgentImageConfig{Tag: "7.78.0"},
+					}).
+					Build()
+				dda.Spec.Features.ServiceDiscovery = &v2alpha1.ServiceDiscoveryFeatureConfig{
+					Enabled: ptr.To(false),
+				}
+				_ = c.Create(context.TODO(), dda)
+				return dda
+			},
+			want:    reconcile.Result{RequeueAfter: defaultRequeueDuration},
+			wantErr: false,
+			wantFunc: func(t *testing.T, c client.Client) {
+				expectedContainers := []string{
+					string(apicommon.CoreAgentContainerName),
+					string(apicommon.TraceAgentContainerName),
+				}
+
+				verifyDaemonsetContainers(t, c, resourcesNamespace, dsName, expectedContainers)
+			},
+		},
+		{
+			name: "DatadogAgent with service discovery explicitly enabled keeps system-probe enabled on unsupported node agent version",
+			loadFunc: func(c client.Client) *v2alpha1.DatadogAgent {
+				dda := testutils.NewInitializedDatadogAgentBuilder(resourcesNamespace, resourcesName).
+					WithComponentOverride(v2alpha1.NodeAgentComponentName, v2alpha1.DatadogAgentComponentOverride{
+						Image: &v2alpha1.AgentImageConfig{Tag: "7.77.2"},
+					}).
+					Build()
+				dda.Spec.Features.ServiceDiscovery = &v2alpha1.ServiceDiscoveryFeatureConfig{
+					Enabled: ptr.To(true),
+				}
+				_ = c.Create(context.TODO(), dda)
+				return dda
+			},
+			want:    reconcile.Result{RequeueAfter: defaultRequeueDuration},
+			wantErr: false,
+			wantFunc: func(t *testing.T, c client.Client) {
+				expectedContainers := []string{
+					string(apicommon.CoreAgentContainerName),
+					string(apicommon.TraceAgentContainerName),
+					string(apicommon.SystemProbeContainerName),
+				}
+
+				verifyDaemonsetContainers(t, c, resourcesNamespace, dsName, expectedContainers)
+			},
+		},
 		{
 			name: "DatadogAgent with Private Action Runner enabled on node, create Daemonset with core, trace, and private-action-runner containers",
 			loadFunc: func(c client.Client) *v2alpha1.DatadogAgent {
@@ -1126,6 +1226,19 @@ func getDeploymentContainers(c client.Client, resourcesNamespace, deploymentName
 	return containers
 }
 
+func assertContainerHasEnvVar(t *testing.T, container corev1.Container, name, value string) {
+	t.Helper()
+
+	for _, envVar := range container.Env {
+		if envVar.Name == name {
+			assert.Equal(t, value, envVar.Value)
+			return
+		}
+	}
+
+	t.Fatalf("env var %s not found in container %s", name, container.Name)
+}
+
 func Test_AutopilotOverrides(t *testing.T) {
 	const resourcesName, resourcesNamespace, dsName = "foo", "bar", "foo-agent"
 
@@ -1827,6 +1940,55 @@ func Test_DDAI_ReconcileV3(t *testing.T) {
 				verifyDDAI(t, c, []v1alpha1.DatadogAgentInternal{*expectedDDAI})
 			},
 		},
+		{
+			name:        "[ddai] Auto-enable service discovery is propagated to DDAI on supported node agent version",
+			ddaiEnabled: true,
+			clientBuilder: fake.NewClientBuilder().
+				WithStatusSubresource(&v2alpha1.DatadogAgent{}, &v1alpha1.DatadogAgentProfile{}, &v1alpha1.DatadogAgentInternal{}),
+			loadFunc: func(c client.Client) *v2alpha1.DatadogAgent {
+				ddaSD := testutils.NewInitializedDatadogAgentBuilder(resourcesNamespace, resourcesName).
+					WithComponentOverride(v2alpha1.NodeAgentComponentName, v2alpha1.DatadogAgentComponentOverride{
+						Image: &v2alpha1.AgentImageConfig{Tag: "7.78.0"},
+					}).
+					Build()
+				_ = c.Create(context.TODO(), ddaSD)
+				return ddaSD
+			},
+			want:    reconcile.Result{RequeueAfter: defaultRequeueDuration},
+			wantErr: false,
+			wantFunc: func(t *testing.T, c client.Client) {
+				ddai := getSingleDDAI(t, c)
+				assert.NotNil(t, ddai.Spec.Features.ServiceDiscovery)
+				assert.NotNil(t, ddai.Spec.Features.ServiceDiscovery.Enabled)
+				assert.True(t, *ddai.Spec.Features.ServiceDiscovery.Enabled)
+			},
+		},
+		{
+			name:        "[ddai] Explicitly disabled service discovery remains disabled in DDAI",
+			ddaiEnabled: true,
+			clientBuilder: fake.NewClientBuilder().
+				WithStatusSubresource(&v2alpha1.DatadogAgent{}, &v1alpha1.DatadogAgentProfile{}, &v1alpha1.DatadogAgentInternal{}),
+			loadFunc: func(c client.Client) *v2alpha1.DatadogAgent {
+				ddaSD := testutils.NewInitializedDatadogAgentBuilder(resourcesNamespace, resourcesName).
+					WithComponentOverride(v2alpha1.NodeAgentComponentName, v2alpha1.DatadogAgentComponentOverride{
+						Image: &v2alpha1.AgentImageConfig{Tag: "7.78.0"},
+					}).
+					Build()
+				ddaSD.Spec.Features.ServiceDiscovery = &v2alpha1.ServiceDiscoveryFeatureConfig{
+					Enabled: ptr.To(false),
+				}
+				_ = c.Create(context.TODO(), ddaSD)
+				return ddaSD
+			},
+			want:    reconcile.Result{RequeueAfter: defaultRequeueDuration},
+			wantErr: false,
+			wantFunc: func(t *testing.T, c client.Client) {
+				ddai := getSingleDDAI(t, c)
+				assert.NotNil(t, ddai.Spec.Features.ServiceDiscovery)
+				assert.NotNil(t, ddai.Spec.Features.ServiceDiscovery.Enabled)
+				assert.False(t, *ddai.Spec.Features.ServiceDiscovery.Enabled)
+			},
+		},
 		{
 			name:        "[ddai] Create DDAI from minimal DDA and default profile",
 			ddaiEnabled: true,
@@ -2040,6 +2202,17 @@ func verifyDDAI(t *testing.T, c client.Client, expectedDDAI []v1alpha1.DatadogAg
 	assert.ElementsMatch(t, expectedDDAI, ddaiList.Items, "DDAI resources don't match")
 }
 
+func getSingleDDAI(t *testing.T, c client.Client) v1alpha1.DatadogAgentInternal {
+	t.Helper()
+
+	ddaiList := v1alpha1.DatadogAgentInternalList{}
+	err := c.List(context.TODO(), &ddaiList)
+	assert.NoError(t, err, "Failed to list DatadogAgentInternal resources")
+	assert.Len(t, ddaiList.Items, 1, "Expected exactly one DatadogAgentInternal resource")
+
+	return ddaiList.Items[0]
+}
+
 func getBaseDDAI(dda *v2alpha1.DatadogAgent) v1alpha1.DatadogAgentInternal {
 	expectedDDAI := v1alpha1.DatadogAgentInternal{
 		ObjectMeta: metav1.ObjectMeta{

internal/controller/datadogagent/defaults/datadogagent_default.go

@@ -14,6 +14,7 @@ import (
 	apiutils "github.com/DataDog/datadog-operator/api/utils"
 	"github.com/DataDog/datadog-operator/internal/controller/datadogagent/common"
 	"github.com/DataDog/datadog-operator/pkg/images"
+	pkgutils "github.com/DataDog/datadog-operator/pkg/utils"
 )
 
 // Default configuration values. These are the recommended settings for monitoring with Datadog in Kubernetes.
@@ -43,7 +44,8 @@ const (
 
 	defaultGPUMonitoringEnabled bool = false
 
-	defaultServiceDiscoveryEnabled bool = false
+	defaultServiceDiscoveryEnabled       bool = false
+	serviceDiscoveryAutoEnableMinVersion      = "7.78.0-0"
 
 	defaultAPMEnabled                   bool   = true
 	defaultAPMHostPortEnabled           bool   = false
@@ -304,7 +306,9 @@ func defaultFeaturesConfig(ddaSpec *v2alpha1.DatadogAgentSpec) {
 	if ddaSpec.Features.ServiceDiscovery == nil {
 		ddaSpec.Features.ServiceDiscovery = &v2alpha1.ServiceDiscoveryFeatureConfig{}
 	}
-	apiutils.DefaultBooleanIfUnset(&ddaSpec.Features.ServiceDiscovery.Enabled, defaultServiceDiscoveryEnabled)
+	if ddaSpec.Features.ServiceDiscovery.Enabled == nil {
+		ddaSpec.Features.ServiceDiscovery.Enabled = ptr.To(shouldEnableServiceDiscoveryByDefault(ddaSpec))
+	}
 
 	// GPU monitoring feature
 	if ddaSpec.Features.GPU == nil {
@@ -648,3 +652,15 @@ func defaultFeaturesConfig(ddaSpec *v2alpha1.DatadogAgentSpec) {
 	}
 	apiutils.DefaultBooleanIfUnset(&ddaSpec.Features.DataPlane.Dogstatsd.Enabled, defaultDataPlaneDogstatsdEnabled)
 }
+
+func shouldEnableServiceDiscoveryByDefault(ddaSpec *v2alpha1.DatadogAgentSpec) bool {
+	fallback := ptr.To(true)
+
+	if ddaSpec.Override != nil {
+		if nodeAgent, found := ddaSpec.Override[v2alpha1.NodeAgentComponentName]; found && nodeAgent != nil && nodeAgent.Image != nil {
+			return pkgutils.IsAboveMinVersion(common.GetAgentVersionFromImage(*nodeAgent.Image), serviceDiscoveryAutoEnableMinVersion, fallback)
+		}
+	}
+
+	return pkgutils.IsAboveMinVersion(images.AgentLatestVersion, serviceDiscoveryAutoEnableMinVersion, fallback)
+}