Cherry-Pick the changes of PCP-3333 newfilter

go.mod

@@ -2,15 +2,15 @@ module sigs.k8s.io/cluster-api
 
 go 1.22
 
-toolchain go1.22.2
+toolchain go1.22.4
 
 require (
 	github.com/MakeNowJust/heredoc v1.0.0
 	github.com/Masterminds/sprig/v3 v3.2.3
 	github.com/blang/semver v3.5.1+incompatible
-	github.com/coredns/corefile-migration v1.0.20
+	github.com/coredns/corefile-migration v1.0.21
 	github.com/davecgh/go-spew v1.1.1
-	github.com/docker/distribution v2.8.1+incompatible
+	github.com/docker/distribution v2.8.2-beta.1+incompatible
 	github.com/drone/envsubst/v2 v2.0.0-20210730161058-179042472c46
 	github.com/evanphx/json-patch/v5 v5.6.0
 	github.com/fatih/color v1.13.0
@@ -116,7 +116,7 @@ require (
 	github.com/prometheus/common v0.37.0 // indirect
 	github.com/prometheus/procfs v0.8.0 // indirect
 	github.com/rivo/uniseg v0.4.2 // indirect
-	github.com/russross/blackfriday v1.5.2 // indirect
+	github.com/russross/blackfriday v1.6.0 // indirect
 	github.com/shopspring/decimal v1.3.1 // indirect
 	github.com/spf13/afero v1.9.2 // indirect
 	github.com/spf13/cast v1.5.0 // indirect

go.sum

@@ -122,8 +122,8 @@ github.com/cncf/udpa/go v0.0.0-20201120205902-5459f2c99403/go.mod h1:WmhPx2Nbnht
 github.com/cncf/xds/go v0.0.0-20210312221358-fbca930ec8ed/go.mod h1:eXthEFrGJvWHgFFCl3hGmgk+/aYT6PnTQLykKQRLhEs=
 github.com/coredns/caddy v1.1.0 h1:ezvsPrT/tA/7pYDBZxu0cT0VmWk75AfIaf6GSYCNMf0=
 github.com/coredns/caddy v1.1.0/go.mod h1:A6ntJQlAWuQfFlsd9hvigKbo2WS0VUs2l1e2F+BawD4=
-github.com/coredns/corefile-migration v1.0.20 h1:MdOkT6F3ehju/n9tgxlGct8XAajOX2vN+wG7To4BWSI=
-github.com/coredns/corefile-migration v1.0.20/go.mod h1:XnhgULOEouimnzgn0t4WPuFDN2/PJQcTxdWKC5eXNGE=
+github.com/coredns/corefile-migration v1.0.21 h1:W/DCETrHDiFo0Wj03EyMkaQ9fwsmSgqTCQDHpceaSsE=
+github.com/coredns/corefile-migration v1.0.21/go.mod h1:XnhgULOEouimnzgn0t4WPuFDN2/PJQcTxdWKC5eXNGE=
 github.com/coreos/bbolt v1.3.2/go.mod h1:iRUV2dpdMOn7Bo10OQBFzIJO9kkE559Wcmn+qkEiiKk=
 github.com/coreos/etcd v3.3.13+incompatible/go.mod h1:uF7uidLiAD3TWHmW31ZFd/JWoc32PjwdhPthX9715RE=
 github.com/coreos/go-semver v0.1.0/go.mod h1:nnelYz7RCh+5ahJtPPxZlU+153eP4D4r3EedlOD2RNk=
@@ -145,8 +145,8 @@ github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/dgrijalva/jwt-go v3.2.0+incompatible/go.mod h1:E3ru+11k8xSBh+hMPgOLZmtrrCbhqsmaPHjLKYnJCaQ=
 github.com/dgryski/go-sip13 v0.0.0-20181026042036-e10d5fee7954/go.mod h1:vAd38F8PWV+bWy6jNmig1y/TA+kYO4g3RSRF0IAv0no=
-github.com/docker/distribution v2.8.1+incompatible h1:Q50tZOPR6T/hjNsyc9g8/syEs6bk8XXApsHjKukMl68=
-github.com/docker/distribution v2.8.1+incompatible/go.mod h1:J2gT2udsDAN96Uj4KfcMRqY0/ypR+oyYUYmja8H+y+w=
+github.com/docker/distribution v2.8.2-beta.1+incompatible h1:gILO60VLD2v28ozemv4aAwDb8ds5U2O/vD/sBXbd7Rw=
+github.com/docker/distribution v2.8.2-beta.1+incompatible/go.mod h1:J2gT2udsDAN96Uj4KfcMRqY0/ypR+oyYUYmja8H+y+w=
 github.com/docopt/docopt-go v0.0.0-20180111231733-ee0de3bc6815/go.mod h1:WwZ+bS3ebgob9U8Nd0kOddGdZWjyMGR8Wziv+TBNwSE=
 github.com/drone/envsubst/v2 v2.0.0-20210730161058-179042472c46 h1:7QPwrLT79GlD5sizHf27aoY2RTvw62mO6x7mxkScNk0=
 github.com/drone/envsubst/v2 v2.0.0-20210730161058-179042472c46/go.mod h1:esf2rsHFNlZlxsqsZDojNBcnNs5REqIvRrWRHqX0vEU=
@@ -509,8 +509,8 @@ github.com/rogpeppe/fastuuid v1.2.0/go.mod h1:jVj6XXZzXRy/MSR5jhDC/2q6DgLz+nrA6L
 github.com/rogpeppe/go-internal v1.3.0/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
 github.com/rogpeppe/go-internal v1.6.1 h1:/FiVV8dS/e+YqF2JvO3yXRFbBLTIuSDkuC7aBOAvL+k=
 github.com/rogpeppe/go-internal v1.6.1/go.mod h1:xXDCJY+GAPziupqXw64V24skbSoqbTEfhy4qGm1nDQc=
-github.com/russross/blackfriday v1.5.2 h1:HyvC0ARfnZBqnXwABFeSZHpKvJHJJfPz81GNueLj0oo=
-github.com/russross/blackfriday v1.5.2/go.mod h1:JO/DiYxRf+HjHt06OyowR9PTA263kcR/rfWxYHBV53g=
+github.com/russross/blackfriday v1.6.0 h1:KqfZb0pUVN2lYqZUYRddxF4OR8ZMURnJIG5Y3VRLtww=
+github.com/russross/blackfriday v1.6.0/go.mod h1:ti0ldHuxg49ri4ksnFxlkCfN+hvslNlmVHqNRXXJNAY=
 github.com/russross/blackfriday/v2 v2.0.1/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
 github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
 github.com/rwcarlsen/goexif v0.0.0-20190401172101-9e8deecbddbd/go.mod h1:hPqNNc0+uJM6H+SuU8sEs5K5IQeKccPqeSjfgcKGgPk=

internal/controllers/machine/machine_controller.go

@@ -57,7 +57,8 @@ import (
 
 const (
 	// controllerName defines the controller used when creating clients.
-	controllerName = "machine-controller"
+	controllerName     = "machine-controller"
+	nodeUnreachableKey = "node.kubernetes.io/unschedulable"
 )
 
 var (
@@ -66,6 +67,11 @@ var (
 	errNoControlPlaneNodes        = errors.New("no control plane members")
 	errClusterIsBeingDeleted      = errors.New("cluster is being deleted")
 	errControlPlaneIsBeingDeleted = errors.New("control plane is being deleted")
+	unreachableToleration         = corev1.Toleration{
+		Key:      nodeUnreachableKey,
+		Effect:   corev1.TaintEffectNoSchedule,
+		Operator: corev1.TolerationOpExists,
+	}
 )
 
 // +kubebuilder:rbac:groups=core,resources=events,verbs=get;list;watch;create;patch
@@ -618,6 +624,9 @@ func (r *Reconciler) drainNode(ctx context.Context, cluster *clusterv1.Cluster,
 		ErrOut: writer{func(msg string, keysAndValues ...interface{}) {
 			log.Error(nil, msg, keysAndValues...)
 		}},
+		AdditionalFilters: []kubedrain.PodFilter{
+			skipUnreachableTolerationPods,
+		},
 		// SPECTRO: Even if the node is reachable, we wait 30 minutes for drain completion else move ahead
 		SkipWaitForDeleteTimeoutSeconds: 60 * 30, // 30 minutes
 	}
@@ -643,6 +652,16 @@ func (r *Reconciler) drainNode(ctx context.Context, cluster *clusterv1.Cluster,
 	return ctrl.Result{}, nil
 }
 
+func skipUnreachableTolerationPods(pod corev1.Pod) kubedrain.PodDeleteStatus {
+	if pod.Spec.Tolerations == nil {
+		return kubedrain.MakePodDeleteStatusOkay()
+	}
+	if HasTolerations(&pod, &unreachableToleration) {
+		return kubedrain.MakePodDeleteStatusSkip()
+	}
+	return kubedrain.MakePodDeleteStatusOkay()
+}
+
 // shouldWaitForNodeVolumes returns true if node status still have volumes attached
 // pod deletion and volume detach happen asynchronously, so pod could be deleted before volume detached from the node
 // this could cause issue for some storage provisioner, for example, vsphere-volume this is problematic

internal/controllers/machine/machine_helpers.go

@@ -17,6 +17,7 @@ limitations under the License.
 package machine
 
 import (
+	corev1 "k8s.io/api/core/v1"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	"k8s.io/apimachinery/pkg/labels"
 )
@@ -37,3 +38,12 @@ func HasMatchingLabels(matchSelector metav1.LabelSelector, matchLabels map[strin
 	}
 	return true
 }
+
+func HasTolerations(pod *corev1.Pod, toleration *corev1.Toleration) bool {
+	for _, t := range pod.Spec.Tolerations {
+		if t.MatchToleration(toleration) {
+			return true
+		}
+	}
+	return false
+}