Bug 1872906: pkg/start: Release leader lease on graceful shutdown

bootstrap/bootstrap-pod.yaml

@@ -37,6 +37,7 @@ spec:
           fieldRef:
             fieldPath: spec.nodeName
   hostNetwork: true
+  terminationGracePeriodSeconds: 130
   volumes:
     - name: kubeconfig
       hostPath:

cmd/start.go

@@ -1,6 +1,8 @@
 package main
 
 import (
+	"context"
+
 	"github.com/spf13/cobra"
 	"k8s.io/klog"
 
@@ -16,11 +18,12 @@ func init() {
 		Long:  "",
 		Run: func(cmd *cobra.Command, args []string) {
 			// To help debugging, immediately log version
-			klog.Infof("%s", version.String)
+			klog.Info(version.String)
 
-			if err := opts.Run(); err != nil {
+			if err := opts.Run(context.Background()); err != nil {
 				klog.Fatalf("error: %v", err)
 			}
+			klog.Infof("Graceful shutdown complete for %s.", version.String)
 		},
 	}
 

docs/user/reconciliation.md

@@ -93,22 +93,22 @@ So the graph nodes are all parallelized with the by-number ordering flattened ou
 
 For the usual reconciliation loop (neither an upgrade between releases nor a fresh install), the flattened graph is also randomly permuted to avoid hanging on ordering bugs.
 
-## Synchronizing the graph
+## Reconciling the graph
 
 The cluster-version operator spawns worker goroutines that walk the graph, pushing manifests in their queue.
-For each manifest in the node, the worker synchronizes the cluster with the manifest using a resource builder.
+For each manifest in the node, the worker reconciles the cluster with the manifest using a resource builder.
 On error (or timeout), the worker abandons the manifest, graph node, and any dependencies of that graph node.
 On success, the worker proceeds to the next manifest in the graph node.
 
 ## Resource builders
 
-Resource builders synchronize the cluster with a manifest from the release image.
+Resource builders reconcile a cluster object with a manifest from the release image.
 The general approach is to generates a merged manifest combining critical spec properties from the release-image manifest with data from a preexisting in-cluster object, if any.
 If the merged manifest differs from the in-cluster object, the merged manifest is pushed back into the cluster.
 
 Some types have additional logic, as described in the following subsections.
 Note that this logic only applies to manifests included in the release image itself.
-For example, only [ClusterOperator](../dev/clusteroperator.md) from the release image will have the blocking logic described [below](#clusteroperator); if an admin or secondary operator pushed a ClusterOperator object, it would not impact the cluster-version operator's graph synchronization.
+For example, only [ClusterOperator](../dev/clusteroperator.md) from the release image will have the blocking logic described [below](#clusteroperator); if an admin or secondary operator pushed a ClusterOperator object, it would not impact the cluster-version operator's graph reconciliation.
 
 ### ClusterOperator
 

docs/user/status.md

@@ -3,6 +3,27 @@
 [The ClusterVersion object](../dev/clusterversion.md) sets `conditions` describing the state of the cluster-version operator (CVO).
 This document describes those conditions and, where appropriate, suggests possible mitigations.
 
+## Failing
+
+When `Failing` is True, the CVO is failing to reconcile the cluster with the desired release image.
+In all cases, the impact on the cluster will be that dependent nodes in [the manifest graph](reconciliation.md#manifest-graph) may not be [reconciled](reconciliation.md#reconciling-the-graph).
+Note that the graph [may be flattened](reconciliation.md#manifest-graph), in which case there are no dependent nodes.
+
+Most reconciliation errors will result in `Failing=True`, although [`ClusterOperatorNotAvailable`](#clusteroperatornotavailable) has special handling.
+
+### NoDesiredImage
+
+The CVO has not been given a release image to reconcile.
+
+If this happens it is a CVO coding error, because clearing [`desiredUpdate`][api-desired-update] should return you to the current CVO's release image.
+
+### ClusterOperatorNotAvailable
+
+`ClusterOperatorNotAvailable` (or the consolidated `ClusterOperatorsNotAvailable`) is set when the CVO fails to retrieve the ClusterOperator from the cluster or when the retrieved ClusterOperator does not satisfy [the reconciliation conditions](reconciliation.md#clusteroperator).
+
+Unlike most manifest-reconciliation failures, this error does not immediately result in `Failing=True`.
+Under some conditions during installs and updates, the CVO will treat this condition as a `Progressing=True` condition and give the operator up to twenty minutes to level before reporting `Failing=True`.
+
 ## RetrievedUpdates
 
 When `RetrievedUpdates` is `True`, the CVO is succesfully retrieving updates, which is good.
@@ -107,5 +128,6 @@ If this error occurs because you forced an update to a release that is not in an
 If this happens it is a CVO coding error.
 There is no mitigation short of updating to a new release image with a fixed CVO.
 
+[api-desired-update]: https://github.com/openshift/api/blob/34f54f12813aaed8822bb5bc56e97cbbfa92171d/config/v1/types_cluster_version.go#L40-L54
 [channels]: https://docs.openshift.com/container-platform/4.3/updating/updating-cluster-between-minor.html#understanding-upgrade-channels_updating-cluster-between-minor
 [Cincinnati]: https://github.com/openshift/cincinnati/blob/master/docs/design/openshift.md

go.mod

@@ -2,6 +2,8 @@ module github.com/openshift/cluster-version-operator
 
 go 1.13
 
+replace golang.org/x/text => golang.org/x/text v0.3.3
+
 require (
 	github.com/blang/semver v3.5.0+incompatible
 	github.com/davecgh/go-spew v1.1.1

go.sum

@@ -397,11 +397,8 @@ golang.org/x/sys v0.0.0-20190801041406-cbf593c0f2f3/go.mod h1:h1NjWce9XRLGQEsW7w
 golang.org/x/sys v0.0.0-20190826190057-c7b8b68b1456/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20191003212358-c178f38b412c h1:6Zx7DRlKXf79yfxuQ/7GqV3w2y7aDsk6bGg0MzF5RVU=
 golang.org/x/sys v0.0.0-20191003212358-c178f38b412c/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
-golang.org/x/text v0.0.0-20160726164857-2910a502d2bf/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
-golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
-golang.org/x/text v0.3.1-0.20180807135948-17ff2d5776d2/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
-golang.org/x/text v0.3.2 h1:tW2bmiBqwgJj/UpqtC8EpXEZVYOwU0yG4iWbprSVAcs=
-golang.org/x/text v0.3.2/go.mod h1:bEr9sfX3Q8Zfm5fL9x+3itogRgK3+ptLWKqgva+5dAk=
+golang.org/x/text v0.3.3 h1:cokOdA+Jmi5PJGXLlLllQSgYigAEfHXJAERHVMaCc2k=
+golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/time v0.0.0-20180412165947-fbb02b2291d2/go.mod h1:tRJNPiyCQ0inRvYxbN9jk5I+vvW/OXSQhTDSoE431IQ=
 golang.org/x/time v0.0.0-20181108054448-85acf8d2951c/go.mod h1:tRJNPiyCQ0inRvYxbN9jk5I+vvW/OXSQhTDSoE431IQ=
 golang.org/x/time v0.0.0-20190308202827-9d24e82272b4/go.mod h1:tRJNPiyCQ0inRvYxbN9jk5I+vvW/OXSQhTDSoE431IQ=

install/0000_00_cluster-version-operator_03_deployment.yaml

@@ -52,6 +52,7 @@ spec:
       nodeSelector:
         node-role.kubernetes.io/master: ""
       priorityClassName: "system-cluster-critical"
+      terminationGracePeriodSeconds: 130
       tolerations:
       - key: "node-role.kubernetes.io/master"
         operator: Exists

lib/resourcemerge/core.go

@@ -164,6 +164,10 @@ func ensureProbePtr(modified *bool, existing **corev1.Probe, required *corev1.Pr
 
 func ensureProbe(modified *bool, existing *corev1.Probe, required corev1.Probe) {
 	setInt32(modified, &existing.InitialDelaySeconds, required.InitialDelaySeconds)
+	setInt32(modified, &existing.TimeoutSeconds, required.TimeoutSeconds)
+	setInt32(modified, &existing.PeriodSeconds, required.PeriodSeconds)
+	setInt32(modified, &existing.SuccessThreshold, required.SuccessThreshold)
+	setInt32(modified, &existing.FailureThreshold, required.FailureThreshold)
 
 	ensureProbeHandler(modified, &existing.Handler, required.Handler)
 }

lib/resourcemerge/core_test.go

@@ -359,6 +359,98 @@ func TestEnsurePodSpec(t *testing.T) {
 				},
 			},
 		},
+		{
+			name: "modify container readiness probe",
+			existing: corev1.PodSpec{
+				Containers: []corev1.Container{
+					{
+						Name: "test",
+						ReadinessProbe: &corev1.Probe{
+							InitialDelaySeconds: 1,
+							TimeoutSeconds:      2,
+							PeriodSeconds:       3,
+							SuccessThreshold:    4,
+							FailureThreshold:    5,
+						},
+					},
+				},
+			},
+			input: corev1.PodSpec{
+				Containers: []corev1.Container{
+					{
+						Name: "test",
+						ReadinessProbe: &corev1.Probe{
+							InitialDelaySeconds: 7,
+							TimeoutSeconds:      8,
+							PeriodSeconds:       9,
+							SuccessThreshold:    10,
+							FailureThreshold:    11,
+						},
+					},
+				},
+			},
+			expectedModified: true,
+			expected: corev1.PodSpec{
+				Containers: []corev1.Container{
+					{
+						Name: "test",
+						ReadinessProbe: &corev1.Probe{
+							InitialDelaySeconds: 7,
+							TimeoutSeconds:      8,
+							PeriodSeconds:       9,
+							SuccessThreshold:    10,
+							FailureThreshold:    11,
+						},
+					},
+				},
+			},
+		},
+		{
+			name: "modify container liveness probe",
+			existing: corev1.PodSpec{
+				Containers: []corev1.Container{
+					{
+						Name: "test",
+						LivenessProbe: &corev1.Probe{
+							InitialDelaySeconds: 1,
+							TimeoutSeconds:      2,
+							PeriodSeconds:       3,
+							SuccessThreshold:    4,
+							FailureThreshold:    5,
+						},
+					},
+				},
+			},
+			input: corev1.PodSpec{
+				Containers: []corev1.Container{
+					{
+						Name: "test",
+						LivenessProbe: &corev1.Probe{
+							InitialDelaySeconds: 7,
+							TimeoutSeconds:      8,
+							PeriodSeconds:       9,
+							SuccessThreshold:    10,
+							FailureThreshold:    11,
+						},
+					},
+				},
+			},
+			expectedModified: true,
+			expected: corev1.PodSpec{
+				Containers: []corev1.Container{
+					{
+						Name: "test",
+						LivenessProbe: &corev1.Probe{
+							InitialDelaySeconds: 7,
+							TimeoutSeconds:      8,
+							PeriodSeconds:       9,
+							SuccessThreshold:    10,
+							FailureThreshold:    11,
+						},
+					},
+				},
+			},
+		},
 	}
 
 	for _, test := range tests {

pkg/autoupdate/autoupdate.go

@@ -87,23 +87,24 @@ func New(
 }
 
 // Run runs the autoupdate controller.
-func (ctrl *Controller) Run(workers int, stopCh <-chan struct{}) {
+func (ctrl *Controller) Run(workers int, stopCh <-chan struct{}) error {
 	defer utilruntime.HandleCrash()
 	defer ctrl.queue.ShutDown()
 
 	klog.Info("Starting AutoUpdateController")
 	defer klog.Info("Shutting down AutoUpdateController")
 
 	if !cache.WaitForCacheSync(stopCh, ctrl.cacheSynced...) {
-		klog.Info("Caches never synchronized")
-		return
+		return fmt.Errorf("caches never synchronized")
 	}
 
 	for i := 0; i < workers; i++ {
+		// FIXME: actually wait until these complete if the Context is canceled.  And possibly add utilruntime.HandleCrash.
 		go wait.Until(ctrl.worker, time.Second, stopCh)
 	}
 
 	<-stopCh
+	return nil
 }
 
 func (ctrl *Controller) eventHandler() cache.ResourceEventHandler {

pkg/cvo/availableupdates.go

@@ -2,7 +2,6 @@ package cvo
 
 import (
 	"crypto/tls"
-	"crypto/x509"
 	"fmt"
 	"net/url"
 	"runtime"
@@ -11,7 +10,6 @@ import (
 	"github.com/blang/semver"
 	"github.com/google/uuid"
 	"k8s.io/apimachinery/pkg/api/equality"
-	"k8s.io/apimachinery/pkg/api/errors"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	"k8s.io/klog"
 
@@ -197,54 +195,3 @@ func calculateAvailableUpdatesStatus(clusterID string, proxyURL *url.URL, tlsCon
 		LastTransitionTime: metav1.Now(),
 	}
 }
-
-// getHTTPSProxyURL returns a url.URL object for the configured
-// https proxy only. It can be nil if does not exist or there is an error.
-func (optr *Operator) getHTTPSProxyURL() (*url.URL, string, error) {
-	proxy, err := optr.proxyLister.Get("cluster")
-
-	if errors.IsNotFound(err) {
-		return nil, "", nil
-	}
-	if err != nil {
-		return nil, "", err
-	}
-
-	if &proxy.Spec != nil {
-		if proxy.Spec.HTTPSProxy != "" {
-			proxyURL, err := url.Parse(proxy.Spec.HTTPSProxy)
-			if err != nil {
-				return nil, "", err
-			}
-			return proxyURL, proxy.Spec.TrustedCA.Name, nil
-		}
-	}
-	return nil, "", nil
-}
-
-func (optr *Operator) getTLSConfig(cmNameRef string) (*tls.Config, error) {
-	cm, err := optr.cmConfigLister.Get(cmNameRef)
-
-	if err != nil {
-		return nil, err
-	}
-
-	certPool, _ := x509.SystemCertPool()
-	if certPool == nil {
-		certPool = x509.NewCertPool()
-	}
-
-	if cm.Data["ca-bundle.crt"] != "" {
-		if ok := certPool.AppendCertsFromPEM([]byte(cm.Data["ca-bundle.crt"])); !ok {
-			return nil, fmt.Errorf("unable to add ca-bundle.crt certificates")
-		}
-	} else {
-		return nil, nil
-	}
-
-	config := &tls.Config{
-		RootCAs: certPool,
-	}
-
-	return config, nil
-}

pkg/cvo/cvo.go

@@ -169,7 +169,6 @@ func New(
 	proxyInformer configinformersv1.ProxyInformer,
 	client clientset.Interface,
 	kubeClient kubernetes.Interface,
-	enableMetrics bool,
 	exclude string,
 ) *Operator {
 	eventBroadcaster := record.NewBroadcaster()
@@ -214,11 +213,6 @@ func New(
 	// make sure this is initialized after all the listers are initialized
 	optr.upgradeableChecks = optr.defaultUpgradeableChecks()
 
-	if enableMetrics {
-		if err := optr.registerMetrics(coInformer.Informer()); err != nil {
-			panic(err)
-		}
-	}
 	return optr
 }
 
@@ -321,8 +315,7 @@ func loadConfigMapVerifierDataFromUpdate(update *payload.Update, clientBuilder v
 }
 
 // Run runs the cluster version operator until stopCh is completed. Workers is ignored for now.
-func (optr *Operator) Run(ctx context.Context, workers int) {
-	defer utilruntime.HandleCrash()
+func (optr *Operator) Run(ctx context.Context, workers int) error {
 	defer optr.queue.ShutDown()
 	stopCh := ctx.Done()
 	workerStopCh := make(chan struct{})
@@ -331,8 +324,7 @@ func (optr *Operator) Run(ctx context.Context, workers int) {
 	defer klog.Info("Shutting down ClusterVersionOperator")
 
 	if !cache.WaitForCacheSync(stopCh, optr.cacheSynced...) {
-		klog.Info("Caches never synchronized")
-		return
+		return fmt.Errorf("caches never synchronized: %w", ctx.Err())
 	}
 
 	// trigger the first cluster version reconcile always
@@ -361,6 +353,8 @@ func (optr *Operator) Run(ctx context.Context, workers int) {
 	// stop the queue, then wait for the worker to exit
 	optr.queue.ShutDown()
 	<-workerStopCh
+
+	return nil
 }
 
 func (optr *Operator) queueKey() string {
@@ -472,7 +466,10 @@ func (optr *Operator) sync(key string) error {
 	// handle the case of a misconfigured CVO by doing nothing
 	if len(desired.Image) == 0 {
 		return optr.syncStatus(original, config, &SyncWorkerStatus{
-			Failure: fmt.Errorf("No configured operator version, unable to update cluster"),
+			Failure: &payload.UpdateError{
+				Reason:  "NoDesiredImage",
+				Message: "No configured operator version, unable to update cluster",
+			},
 		}, errs)
 	}