Allow Agent pods to have Affinity (#319)

api/v1alpha1/datadogagent_types.go

@@ -238,6 +238,10 @@ type DatadogAgentSpecAgentSpec struct {
 	// Provide Agent Network Policy configuration
 	// +optional
 	NetworkPolicy NetworkPolicySpec `json:"networkPolicy,omitempty"`
+
+	// If specified, the pod's scheduling constraints.
+	// +optional
+	Affinity *corev1.Affinity `json:"affinity,omitempty"`
 }
 
 // RbacConfig contains RBAC configuration.

controllers/datadogagent/controller_test.go

@@ -155,6 +155,20 @@ func TestReconcileDatadogAgent_Reconcile(t *testing.T) {
 	s.AddKnownTypes(networkingv1.SchemeGroupVersion, &networkingv1.NetworkPolicy{})
 
 	defaultRequeueDuration := 15 * time.Second
+	affinity := &corev1.Affinity{
+		PodAntiAffinity: &corev1.PodAntiAffinity{
+			RequiredDuringSchedulingIgnoredDuringExecution: []corev1.PodAffinityTerm{
+				{
+					LabelSelector: &metav1.LabelSelector{
+						MatchLabels: map[string]string{
+							"foo": "bar",
+						},
+					},
+					TopologyKey: "baz",
+				},
+			},
+		},
+	}
 
 	type fields struct {
 		client   client.Client
@@ -2225,6 +2239,41 @@ func TestReconcileDatadogAgent_Reconcile(t *testing.T) {
 				return nil
 			},
 		},
+		{
+			name: "DatadogAgent found and defaulted, DaemonSet has Affinity",
+			fields: fields{
+				client:   fake.NewFakeClient(),
+				scheme:   s,
+				recorder: recorder,
+			},
+			args: args{
+				request: newRequest(resourcesNamespace, resourcesName),
+				loadFunc: func(c client.Client) {
+					dadOptions := &test.NewDatadogAgentOptions{}
+					dda := test.NewDefaultedDatadogAgent(resourcesNamespace, resourcesName, dadOptions)
+
+					dda.Spec.Agent.Affinity = affinity
+
+					_ = c.Create(context.TODO(), dda)
+					createAgentDependencies(c, dda)
+				},
+			},
+			want:    reconcile.Result{RequeueAfter: defaultRequeueDuration},
+			wantErr: false,
+			wantFunc: func(c client.Client) error {
+				ds := &appsv1.DaemonSet{}
+				err := c.Get(context.TODO(), newRequest(resourcesNamespace, dsName).NamespacedName, ds)
+				if err != nil {
+					return err
+				}
+
+				if !reflect.DeepEqual(ds.Spec.Template.Spec.Affinity, affinity) {
+					return fmt.Errorf("pod affinity does not match the one specified. got: %+v", ds.Spec.Template.Spec.Affinity)
+				}
+
+				return nil
+			},
+		},
 		{
 			name: "DatadogAgent found and defaulted, Cluster Agent network policies are created",
 			fields: fields{

controllers/datadogagent/utils.go

@@ -151,6 +151,7 @@ func newAgentPodTemplate(logger logr.Logger, dda *datadoghqv1alpha1.DatadogAgent
 			HostPID:            dda.Spec.Agent.HostPID || isComplianceEnabled(&dda.Spec),
 			DNSPolicy:          dda.Spec.Agent.DNSPolicy,
 			DNSConfig:          dda.Spec.Agent.DNSConfig,
+			Affinity:           dda.Spec.Agent.Affinity,
 		},
 	}, nil
 }

docs/configuration.md

@@ -30,6 +30,12 @@ spec:
 | --------- | ----------- |
 | agent.additionalAnnotations | AdditionalAnnotations provide annotations that will be added to the Agent Pods. |
 | agent.additionalLabels | AdditionalLabels provide labels that will be added to the Agent Pods. |
+| agent.affinity.nodeAffinity.preferredDuringSchedulingIgnoredDuringExecution | The scheduler will prefer to schedule pods to nodes that satisfy the affinity expressions specified by this field, but it may choose a node that violates one or more of the expressions. The node that is most preferred is the one with the greatest sum of weights, i.e. for each node that meets all of the scheduling requirements (resource request, requiredDuringScheduling affinity expressions, etc.), compute a sum by iterating through the elements of this field and adding "weight" to the sum if the node matches the corresponding matchExpressions; the node(s) with the highest sum are the most preferred. |
+| agent.affinity.nodeAffinity.requiredDuringSchedulingIgnoredDuringExecution.nodeSelectorTerms | Required. A list of node selector terms. The terms are ORed. |
+| agent.affinity.podAffinity.preferredDuringSchedulingIgnoredDuringExecution | The scheduler will prefer to schedule pods to nodes that satisfy the affinity expressions specified by this field, but it may choose a node that violates one or more of the expressions. The node that is most preferred is the one with the greatest sum of weights, i.e. for each node that meets all of the scheduling requirements (resource request, requiredDuringScheduling affinity expressions, etc.), compute a sum by iterating through the elements of this field and adding "weight" to the sum if the node has pods which matches the corresponding podAffinityTerm; the node(s) with the highest sum are the most preferred. |
+| agent.affinity.podAffinity.requiredDuringSchedulingIgnoredDuringExecution | If the affinity requirements specified by this field are not met at scheduling time, the pod will not be scheduled onto the node. If the affinity requirements specified by this field cease to be met at some point during pod execution (e.g. due to a pod label update), the system may or may not try to eventually evict the pod from its node. When there are multiple elements, the lists of nodes corresponding to each podAffinityTerm are intersected, i.e. all terms must be satisfied. |
+| agent.affinity.podAntiAffinity.preferredDuringSchedulingIgnoredDuringExecution | The scheduler will prefer to schedule pods to nodes that satisfy the anti-affinity expressions specified by this field, but it may choose a node that violates one or more of the expressions. The node that is most preferred is the one with the greatest sum of weights, i.e. for each node that meets all of the scheduling requirements (resource request, requiredDuringScheduling anti-affinity expressions, etc.), compute a sum by iterating through the elements of this field and adding "weight" to the sum if the node has pods which matches the corresponding podAffinityTerm; the node(s) with the highest sum are the most preferred. |
+| agent.affinity.podAntiAffinity.requiredDuringSchedulingIgnoredDuringExecution | If the anti-affinity requirements specified by this field are not met at scheduling time, the pod will not be scheduled onto the node. If the anti-affinity requirements specified by this field cease to be met at some point during pod execution (e.g. due to a pod label update), the system may or may not try to eventually evict the pod from its node. When there are multiple elements, the lists of nodes corresponding to each podAffinityTerm are intersected, i.e. all terms must be satisfied. |
 | agent.apm.args | Args allows the specification of extra args to `Command` parameter |
 | agent.apm.command | Command allows the specification of custom entrypoint for Trace Agent container |
 | agent.apm.enabled | Enable this to enable APM and tracing, on port 8126. See also: https://github.com/DataDog/docker-dd-agent#tracing-from-the-host |