Set DNS trusted servers using FelixConfig instead of FELIX_ env var

pkg/apis/crd.projectcalico.org/v1/felixconfig.go

@@ -379,6 +379,14 @@ type FelixConfigurationSpec struct {
 	EgressIPVXLANPort *int `json:"egressIPVXLANPort,omitempty"`
 	// EgressIPVXLANVNI is the VNI ID of vxlan tunnel device for egress traffic. [Default: 4097]
 	EgressIPVXLANVNI *int `json:"egressIPVXLANVNI,omitempty"`
+
+	// The DNS servers that Felix should trust. Each entry here must be `<ip>[:<port>]` - indicating an
+	// explicit DNS server IP - or `k8s-service:[<namespace>/]<name>[:port]` - indicating a Kubernetes DNS
+	// service. `<port>` defaults to the first service port, or 53 for an IP, and `<namespace>` to
+	// `kube-system`. An IPv6 address with a port must use the square brackets convention, for example
+	// `[fd00:83a6::12]:5353`.Note that Felix (calico-node) will need RBAC permission to read the details of
+	// each service specified by a `k8s-service:...` form. [Default: "k8s-service:kube-dns"].
+	DNSTrustedServers *[]string `json:"dnsTrustedServers,omitempty"`
 }
 
 type RouteTableRange struct {

pkg/controller/applicationlayer/applicationlayer_controller.go

@@ -521,58 +521,47 @@ func (r *ReconcileApplicationLayer) getTProxyMode(al *operatorv1.ApplicationLaye
 // patchFelixConfiguration takes all application layer specs as arguments and patches felix config.
 // If at least one of the specs requires TPROXYMode as "Enabled" it'll be patched as "Enabled" otherwise it is "Disabled".
 func (r *ReconcileApplicationLayer) patchFelixConfiguration(ctx context.Context, al *operatorv1.ApplicationLayer) error {
-	// Fetch any existing default FelixConfiguration object.
-	fc := &crdv1.FelixConfiguration{}
-	err := r.client.Get(ctx, types.NamespacedName{Name: "default"}, fc)
-	if err != nil && !apierrors.IsNotFound(err) {
-		r.status.SetDegraded(operatorv1.ResourceReadError, "Unable to read FelixConfiguration", err, log)
-		return err
-	}
+	_, err := utils.PatchFelixConfiguration(ctx, r.client, func(fc *crdv1.FelixConfiguration) bool {
+		var tproxyMode crdv1.TPROXYModeOption
+		if ok, v := r.getTProxyMode(al); ok {
+			tproxyMode = v
+		} else {
+			if fc.Spec.TPROXYMode == nil {
+				// Workaround: we'd like to always force the value to be the correct one, matching the operator's
+				// configuration.  However, during an upgrade from a version that predates the TPROXYMode option,
+				// Felix hits a bug and gets confused by the new config parameter, which in turn triggers a restart.
+				// Work around that by relying on Disabled being the default value for the field instead.
+				//
+				// The felix bug was fixed in v3.16, v3.15.1 and v3.14.4; it should be safe to set new config fields
+				// once we know we're only upgrading from those versions and above.
+				return false
+			}
 
-	patchFrom := client.MergeFrom(fc.DeepCopy())
-
-	var tproxyMode crdv1.TPROXYModeOption
-	if ok, v := r.getTProxyMode(al); ok {
-		tproxyMode = v
-	} else {
-		if fc.Spec.TPROXYMode == nil {
-			// Workaround: we'd like to always force the value to be the correct one, matching the operator's
-			// configuration.  However, during an upgrade from a version that predates the TPROXYMode option,
-			// Felix hits a bug and gets confused by the new config parameter, which in turn triggers a restart.
-			// Work around that by relying on Disabled being the default value for the field instead.
-			//
-			// The felix bug was fixed in v3.16, v3.15.1 and v3.14.4; it should be safe to set new config fields
-			// once we know we're only upgrading from those versions and above.
-			return nil
+			// If the mode is already set, fall through to the normal logic, it's safe to force-set the field now.
+			// This also avoids churning the config if a previous version of the operator set it to Disabled already,
+			// we avoid setting it back to nil.
+			tproxyMode = crdv1.TPROXYModeOptionDisabled
 		}
 
-		// If the mode is already set, fall through to the normal logic, it's safe to force-set the field now.
-		// This also avoids churning the config if a previous version of the operator set it to Disabled already,
-		// we avoid setting it back to nil.
-		tproxyMode = crdv1.TPROXYModeOptionDisabled
-	}
-
-	policySyncPrefix := r.getPolicySyncPathPrefix(&fc.Spec, al)
-	policySyncPrefixSetDesired := fc.Spec.PolicySyncPathPrefix == policySyncPrefix
-	tproxyModeSetDesired := fc.Spec.TPROXYMode != nil && *fc.Spec.TPROXYMode == tproxyMode
+		policySyncPrefix := r.getPolicySyncPathPrefix(&fc.Spec, al)
+		policySyncPrefixSetDesired := fc.Spec.PolicySyncPathPrefix == policySyncPrefix
+		tproxyModeSetDesired := fc.Spec.TPROXYMode != nil && *fc.Spec.TPROXYMode == tproxyMode
 
-	// If tproxy mode is already set to desired state return nil.
-	if policySyncPrefixSetDesired && tproxyModeSetDesired {
-		return nil
-	}
-
-	fc.Spec.TPROXYMode = &tproxyMode
-	fc.Spec.PolicySyncPathPrefix = policySyncPrefix
+		// If tproxy mode is already set to desired state return false to indicate patch not needed.
+		if policySyncPrefixSetDesired && tproxyModeSetDesired {
+			return false
+		}
 
-	log.Info(
-		"Patching FelixConfiguration: ",
-		"policySyncPathPrefix", fc.Spec.PolicySyncPathPrefix,
-		"tproxyMode", string(tproxyMode),
-	)
+		fc.Spec.TPROXYMode = &tproxyMode
+		fc.Spec.PolicySyncPathPrefix = policySyncPrefix
 
-	if err := r.client.Patch(ctx, fc, patchFrom); err != nil {
-		return err
-	}
+		log.Info(
+			"Patching FelixConfiguration: ",
+			"policySyncPathPrefix", fc.Spec.PolicySyncPathPrefix,
+			"tproxyMode", string(tproxyMode),
+		)
+		return true
+	})
 
-	return nil
+	return err
 }

pkg/controller/egressgateway/egressgateway_controller.go

@@ -309,7 +309,13 @@ func (r *ReconcileEgressGateway) Reconcile(ctx context.Context, request reconcil
 	}
 
 	// patch and get the felix configuration
-	fc, err := r.patchFelixConfig(ctx)
+	fc, err := utils.PatchFelixConfiguration(ctx, r.client, func(fc *crdv1.FelixConfiguration) bool {
+		if fc.Spec.PolicySyncPathPrefix != "" {
+			return false // don't proceed with the patch
+		}
+		fc.Spec.PolicySyncPathPrefix = "/var/run/nodeagent"
+		return true // proceed with this patch
+	})
 	if err != nil {
 		reqLogger.Error(err, "Error patching felix configuration")
 		r.status.SetDegraded(operatorv1.ResourcePatchError, "Error patching felix configuration", err, reqLogger)
@@ -430,26 +436,6 @@ func (r *ReconcileEgressGateway) reconcileEgressGateway(ctx context.Context, egw
 	return nil
 }
 
-func (r *ReconcileEgressGateway) patchFelixConfig(ctx context.Context) (*crdv1.FelixConfiguration, error) {
-	// Fetch any existing default FelixConfiguration object.
-	fc := &crdv1.FelixConfiguration{}
-	err := r.client.Get(ctx, types.NamespacedName{Name: "default"}, fc)
-	if err != nil && !errors.IsNotFound(err) {
-		r.status.SetDegraded(operatorv1.ResourceReadError, "Unable to read FelixConfiguration", err, log)
-		return nil, err
-	}
-
-	patchFrom := client.MergeFrom(fc.DeepCopy())
-	if fc.Spec.PolicySyncPathPrefix != "" {
-		return fc, nil
-	}
-	fc.Spec.PolicySyncPathPrefix = "/var/run/nodeagent"
-	if err := r.client.Patch(ctx, fc, patchFrom); err != nil {
-		return nil, err
-	}
-	return fc, nil
-}
-
 // getRequestedEgressGateway returns the namespaced EgressGateway instance.
 func getRequestedEgressGateway(egws []operatorv1.EgressGateway, request reconcile.Request) (*operatorv1.EgressGateway, int) {
 	for index, egw := range egws {

pkg/controller/installation/core_controller.go

@@ -1133,15 +1133,11 @@ func (r *ReconcileInstallation) Reconcile(ctx context.Context, request reconcile
 		}
 	}
 
-	// Fetch any existing default FelixConfiguration object.
-	felixConfiguration := &crdv1.FelixConfiguration{}
-	err = r.client.Get(ctx, types.NamespacedName{Name: "default"}, felixConfiguration)
-	if err != nil && !apierrors.IsNotFound(err) {
-		r.status.SetDegraded(operator.ResourceReadError, "Unable to read FelixConfiguration", err, reqLogger)
-		return reconcile.Result{}, err
-	}
-
-	if err = r.setDefaultsOnFelixConfiguration(ctx, instance, felixConfiguration, reqLogger); err != nil {
+	// Set any non-default FelixConfiguration values that we need.
+	felixConfiguration, err := utils.PatchFelixConfiguration(ctx, r.client, func(fc *crdv1.FelixConfiguration) bool {
+		return r.setDefaultsOnFelixConfiguration(instance, fc)
+	})
+	if err != nil {
 		return reconcile.Result{}, err
 	}
 
@@ -1618,11 +1614,8 @@ func GetOrCreateTyphaNodeTLSConfig(cli client.Client, certificateManager certifi
 }
 
 // setDefaultOnFelixConfiguration will take the passed in fc and add any defaulting needed
-// based on the install config. If the FelixConfig ResourceVersion is empty,
-// then the FelixConfig default will be created, otherwise a patch will be performed.
-func (r *ReconcileInstallation) setDefaultsOnFelixConfiguration(ctx context.Context, install *operator.Installation, fc *crdv1.FelixConfiguration, log logr.Logger) error {
-	patchFrom := client.MergeFrom(fc.DeepCopy())
-	fc.ObjectMeta.Name = "default"
+// based on the install config.
+func (r *ReconcileInstallation) setDefaultsOnFelixConfiguration(install *operator.Installation, fc *crdv1.FelixConfiguration) bool {
 	updated := false
 
 	switch install.Spec.CNI.Type {
@@ -1665,21 +1658,37 @@ func (r *ReconcileInstallation) setDefaultsOnFelixConfiguration(ctx context.Cont
 		updated = true
 	}
 
-	if !updated {
-		return nil
-	}
-	if fc.ResourceVersion == "" {
-		if err := r.client.Create(ctx, fc); err != nil {
-			r.status.SetDegraded(operator.ResourceCreateError, "Unable to Create default FelixConfiguration", err, log)
-			return err
-		}
-	} else {
-		if err := r.client.Patch(ctx, fc, patchFrom); err != nil {
-			r.status.SetDegraded(operator.ResourcePatchError, "Unable to Patch default FelixConfiguration", err, log)
-			return err
+	if install.Spec.Variant == operator.TigeraSecureEnterprise {
+		// Some platforms need a different default setting for dnsTrustedServers, because their DNS service is not named "kube-dns".
+		dnsService := ""
+		switch install.Spec.KubernetesProvider {
+		case operator.ProviderOpenShift:
+			dnsService = "k8s-service:openshift-dns/dns-default"
+		case operator.ProviderRKE2:
+			dnsService = "k8s-service:kube-system/rke2-coredns-rke2-coredns"
+		}
+		if dnsService != "" {
+			felixDefault := "k8s-service:kube-dns"
+			trustedServers := []string{dnsService}
+			// Keep any other values that are already configured, excepting the value
+			// that we are setting and the kube-dns default.
+			existingSetting := ""
+			if fc.Spec.DNSTrustedServers != nil {
+				existingSetting = strings.Join(*(fc.Spec.DNSTrustedServers), ",")
+				for _, server := range *(fc.Spec.DNSTrustedServers) {
+					if server != felixDefault && server != dnsService {
+						trustedServers = append(trustedServers, server)
+					}
+				}
+			}
+			newSetting := strings.Join(trustedServers, ",")
+			if newSetting != existingSetting {
+				fc.Spec.DNSTrustedServers = &trustedServers
+				updated = true
+			}
 		}
 	}
-	return nil
+	return updated
 }
 
 var osExitOverride = os.Exit

pkg/controller/installation/core_controller_test.go

@@ -1073,6 +1073,20 @@ var _ = Describe("Testing core-controller installation", func() {
 			Expect(fc.Spec.RouteTableRange).To(BeNil())
 		})
 
+		It("generates FelixConfiguration with correct DNS service for Rancher", func() {
+			cr.Spec.KubernetesProvider = operator.ProviderRKE2
+			Expect(c.Create(ctx, cr)).NotTo(HaveOccurred())
+			_, err := r.Reconcile(ctx, reconcile.Request{})
+			Expect(err).ShouldNot(HaveOccurred())
+
+			// We should get a felix configuration with Rancher's DNS service.
+			fc := &crdv1.FelixConfiguration{}
+			err = c.Get(ctx, types.NamespacedName{Name: "default"}, fc)
+			Expect(err).ShouldNot(HaveOccurred())
+			Expect(fc.Spec.DNSTrustedServers).NotTo(BeNil())
+			Expect(*fc.Spec.DNSTrustedServers).To(ConsistOf("k8s-service:kube-system/rke2-coredns-rke2-coredns"))
+		})
+
 		It("should Reconcile with AWS CNI config", func() {
 			cr.Spec.CNI = &operator.CNISpec{Type: operator.PluginAmazonVPC}
 			Expect(c.Create(ctx, cr)).NotTo(HaveOccurred())

pkg/controller/utils/felix_configuration.go

@@ -0,0 +1,54 @@
+// Copyright (c) 2023 Tigera, Inc. All rights reserved.
+
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package utils
+
+import (
+	"context"
+	"fmt"
+
+	crdv1 "github.com/tigera/operator/pkg/apis/crd.projectcalico.org/v1"
+	"k8s.io/apimachinery/pkg/api/errors"
+	"k8s.io/apimachinery/pkg/types"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+)
+
+func PatchFelixConfiguration(ctx context.Context, c client.Client, patchFn func(fc *crdv1.FelixConfiguration) bool) (*crdv1.FelixConfiguration, error) {
+	// Fetch any existing default FelixConfiguration object.
+	fc := &crdv1.FelixConfiguration{}
+	err := c.Get(ctx, types.NamespacedName{Name: "default"}, fc)
+	if err != nil && !errors.IsNotFound(err) {
+		return nil, fmt.Errorf("unable to read FelixConfiguration: %w", err)
+	}
+
+	// Create a base state for the upcoming patch operation.
+	patchFrom := client.MergeFrom(fc.DeepCopy())
+
+	// Apply desired changes to the FelixConfiguration.
+	if patchFn(fc) {
+		// Apply the patch.
+		if fc.ResourceVersion == "" {
+			fc.ObjectMeta.Name = "default"
+			if err := c.Create(ctx, fc); err != nil {
+				return nil, err
+			}
+		} else {
+			if err := c.Patch(ctx, fc, patchFrom); err != nil {
+				return nil, err
+			}
+		}
+	}
+
+	return fc, nil
+}

pkg/render/node.go

@@ -727,7 +727,7 @@ func (c *nodeComponent) nodeCNIConfigMap() *corev1.ConfigMap {
 	config := fmt.Sprintf(`{
 			  "name": "k8s-pod-network",
 			  "cniVersion": "0.3.1",
-			  "plugins": %s 
+			  "plugins": %s
 			}`, string(pluginsArray))
 
 	return &corev1.ConfigMap{
@@ -1650,15 +1650,6 @@ func (c *nodeComponent) nodeEnvVars() []corev1.EnvVar {
 	case operatorv1.ProviderOpenShift:
 		// For Openshift, we need special configuration since our default port is already in use.
 		nodeEnv = append(nodeEnv, corev1.EnvVar{Name: "FELIX_HEALTHPORT", Value: "9199"})
-		if c.cfg.Installation.Variant == operatorv1.TigeraSecureEnterprise {
-			// We also need to configure a non-default trusted DNS server, since there's no kube-dns.
-			nodeEnv = append(nodeEnv, corev1.EnvVar{Name: "FELIX_DNSTRUSTEDSERVERS", Value: "k8s-service:openshift-dns/dns-default"})
-		}
-	case operatorv1.ProviderRKE2:
-		// For RKE2, configure a non-default trusted DNS server, as the DNS service is not named "kube-dns".
-		if c.cfg.Installation.Variant == operatorv1.TigeraSecureEnterprise {
-			nodeEnv = append(nodeEnv, corev1.EnvVar{Name: "FELIX_DNSTRUSTEDSERVERS", Value: "k8s-service:kube-system/rke2-coredns-rke2-coredns"})
-		}
 	// For AKS/AzureVNET and EKS/VPCCNI, we must explicitly ask felix to add host IP's to wireguard ifaces
 	case operatorv1.ProviderAKS:
 		if c.cfg.Installation.CNI.Type == operatorv1.PluginAzureVNET {

pkg/render/node_test.go

@@ -1951,7 +1951,6 @@ var _ = Describe("Node rendering tests", func() {
 					// The OpenShift envvar overrides.
 					{Name: "FELIX_HEALTHPORT", Value: "9199"},
 					{Name: "MULTI_INTERFACE_MODE", Value: operatorv1.MultiInterfaceModeNone.Value()},
-					{Name: "FELIX_DNSTRUSTEDSERVERS", Value: "k8s-service:openshift-dns/dns-default"},
 					{Name: "FIPS_MODE_ENABLED", Value: "false"},
 				}
 				expectedNodeEnv = configureExpectedNodeEnvIPVersions(expectedNodeEnv, defaultInstance, enableIPv4, enableIPv6)
@@ -2046,7 +2045,6 @@ var _ = Describe("Node rendering tests", func() {
 
 					// The RKE2 envvar overrides.
 					{Name: "MULTI_INTERFACE_MODE", Value: operatorv1.MultiInterfaceModeNone.Value()},
-					{Name: "FELIX_DNSTRUSTEDSERVERS", Value: "k8s-service:kube-system/rke2-coredns-rke2-coredns"},
 					{Name: "FIPS_MODE_ENABLED", Value: "false"},
 				}
 				expectedNodeEnv = configureExpectedNodeEnvIPVersions(expectedNodeEnv, defaultInstance, enableIPv4, enableIPv6)