Add feature gate for disabling the MHC controller #1151
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -2,6 +2,7 @@ package operator | |
|
|
||
| import ( | ||
| "context" | ||
| "errors" | ||
| "fmt" | ||
| "os" | ||
| "time" | ||
|
|
@@ -12,9 +13,10 @@ import ( | |
| configinformersv1 "github.com/openshift/client-go/config/informers/externalversions/config/v1" | ||
| configlistersv1 "github.com/openshift/client-go/config/listers/config/v1" | ||
| machineclientset "github.com/openshift/client-go/machine/clientset/versioned" | ||
| "github.com/openshift/library-go/pkg/operator/configobserver/featuregates" | ||
| "github.com/openshift/library-go/pkg/operator/events" | ||
| "github.com/openshift/library-go/pkg/operator/resource/resourceapply" | ||
| admissionregistrationv1 "k8s.io/api/admissionregistration/v1" | ||
| metav1 "k8s.io/apimachinery/pkg/apis/meta/v1" | ||
| utilruntime "k8s.io/apimachinery/pkg/util/runtime" | ||
| "k8s.io/apimachinery/pkg/util/wait" | ||
|
|
@@ -39,6 +41,9 @@ const ( | |
| // 5ms, 10ms, 20ms, 40ms, 80ms, 160ms, 320ms, 640ms, 1.3s, 2.6s, 5.1s, 10.2s, 20.4s, 41s, 82s | ||
| maxRetries = 15 | ||
| maoOwnedAnnotation = "machine.openshift.io/owned" | ||
|
|
||
| releaseVersionEnvVariableName = "RELEASE_VERSION" | ||
| releaseVersionUnknownValue = "unknown" | ||
|
Contributor
There was a problem hiding this comment. Is this the same value as the raw version string when it's not initialised?
Member
Author
There was a problem hiding this comment. I didn't understand how/when this is used tbh, I copied it from here: https://github.com/openshift/cluster-cloud-controller-manager-operator/blob/43facae32a53c73ad567129150db6865673bdf1d/pkg/controllers/status.go#L224
Contributor
There was a problem hiding this comment. Ok, I think it's fine to leave as that for now, lets see if anything breaks! |
||
| ) | ||
|
|
||
| // Operator defines machine api operator. | ||
|
|
@@ -53,6 +58,7 @@ type Operator struct { | |
| machineClient machineclientset.Interface | ||
| dynamicClient dynamic.Interface | ||
| eventRecorder record.EventRecorder | ||
| recorder events.Recorder | ||
|
|
||
| syncHandler func(ic string) (reconcile.Result, error) | ||
|
|
||
|
|
@@ -71,8 +77,7 @@ type Operator struct { | |
| mutatingWebhookLister admissionlisterv1.MutatingWebhookConfigurationLister | ||
| mutatingWebhookListerSynced cache.InformerSynced | ||
|
|
||
| featureGateAccessor featuregates.FeatureGateAccess | ||
|
|
||
| // queue only ever has one item, but it has nice error handling backoff/retry semantics | ||
| queue workqueue.RateLimitingInterface | ||
|
|
@@ -91,6 +96,7 @@ func New( | |
| deployInformer appsinformersv1.DeploymentInformer, | ||
| daemonsetInformer appsinformersv1.DaemonSetInformer, | ||
| featureGateInformer configinformersv1.FeatureGateInformer, | ||
| clusterVersionInformer configinformersv1.ClusterVersionInformer, | ||
| validatingWebhookInformer admissioninformersv1.ValidatingWebhookConfigurationInformer, | ||
| mutatingWebhookInformer admissioninformersv1.MutatingWebhookConfigurationInformer, | ||
| proxyInformer configinformersv1.ProxyInformer, | ||
|
|
@@ -99,13 +105,18 @@ func New( | |
| machineClient machineclientset.Interface, | ||
| dynamicClient dynamic.Interface, | ||
|
|
||
| eventRecorder record.EventRecorder, | ||
| recorder events.Recorder, | ||
| ) (*Operator, error) { | ||
| // we must report the version from the release payload when we report available at that level | ||
| // TODO we will report the version of the operands (so our machine api implementation version) | ||
| operandVersions := []osconfigv1.OperandVersion{} | ||
| releaseVersion := os.Getenv(releaseVersionEnvVariableName) | ||
| if len(releaseVersion) > 0 { | ||
| operandVersions = append(operandVersions, osconfigv1.OperandVersion{Name: "operator", Version: releaseVersion}) | ||
| } else { | ||
| klog.Infof("%s environment variable is missing, defaulting to %q", releaseVersionEnvVariableName, releaseVersionUnknownValue) | ||
| releaseVersion = releaseVersionUnknownValue | ||
| } | ||
|
|
||
| optr := &Operator{ | ||
|
|
@@ -116,7 +127,8 @@ func New( | |
| osClient: osClient, | ||
| machineClient: machineClient, | ||
| dynamicClient: dynamicClient, | ||
| eventRecorder: eventRecorder, | ||
| recorder: recorder, | ||
| queue: workqueue.NewNamedRateLimitingQueue(workqueue.DefaultControllerRateLimiter(), "machineapioperator"), | ||
| operandVersions: operandVersions, | ||
| } | ||
|
|
@@ -137,10 +149,36 @@ func New( | |
| if err != nil { | ||
| return nil, fmt.Errorf("error adding event handler to mutatingwebhook informer: %v", err) | ||
| } | ||
|
|
||
| desiredVersion := releaseVersion | ||
| missingVersion := "0.0.1-snapshot" | ||
| featureGateAccessor := featuregates.NewFeatureGateAccess( | ||
| desiredVersion, missingVersion, | ||
| clusterVersionInformer, featureGateInformer, | ||
| recorder, | ||
| ) | ||
| featureGateAccessor.SetChangeHandler(func(featureChange featuregates.FeatureChange) { | ||
| if featureChange.Previous == nil { | ||
| // When the initial featuregate is set, the previous version is nil. | ||
| // Nothing to do in this case, it's handled by the 1st sync, which only runs after the initial feature set was received. | ||
| return | ||
| } | ||
|
|
||
| klog.V(4).InfoS("FeatureGates changed", "enabled", featureChange.New.Enabled, "disabled", featureChange.New.Disabled) | ||
| prevDisableMHC := featuregates.NewFeatureGate(featureChange.Previous.Enabled, featureChange.Previous.Disabled). | ||
| Enabled(osconfigv1.FeatureGateMachineAPIOperatorDisableMachineHealthCheckController) | ||
| newDisableMHC := featuregates.NewFeatureGate(featureChange.New.Enabled, featureChange.New.Disabled). | ||
| Enabled(osconfigv1.FeatureGateMachineAPIOperatorDisableMachineHealthCheckController) | ||
|
|
||
| if prevDisableMHC != newDisableMHC { | ||
| klog.V(2).InfoS("Resync for modified feature gate", | ||
| "FeatureGateMachineAPIOperatorDisableMachineHealthCheckController enabled", newDisableMHC, | ||
| ) | ||
| workQueueKey := fmt.Sprintf("%s/%s", optr.namespace, optr.name) | ||
| optr.queue.Add(workQueueKey) | ||
| } | ||
| }) | ||
| optr.featureGateAccessor = featureGateAccessor | ||
|
|
||
| optr.config = config | ||
| optr.syncHandler = optr.sync | ||
|
|
@@ -160,9 +198,6 @@ func New( | |
| optr.mutatingWebhookLister = mutatingWebhookInformer.Lister() | ||
| optr.mutatingWebhookListerSynced = mutatingWebhookInformer.Informer().HasSynced | ||
|
|
||
| return optr, nil | ||
| } | ||
|
|
||
|
|
@@ -179,12 +214,24 @@ func (optr *Operator) Run(workers int, stopCh <-chan struct{}) { | |
| optr.validatingWebhookListerSynced, | ||
| optr.deployListerSynced, | ||
| optr.daemonsetListerSynced, | ||
| optr.proxyListerSynced) { | ||
| klog.Error("Failed to sync caches") | ||
| return | ||
| } | ||
| klog.Info("Synced up caches") | ||
|
|
||
| ctx, cancelFeatureGateAccessor := context.WithCancel(context.Background()) | ||
| defer cancelFeatureGateAccessor() | ||
| go optr.featureGateAccessor.Run(ctx) | ||
| klog.Info("Started feature gate accessor") | ||
| select { | ||
| case <-optr.featureGateAccessor.InitialFeatureGatesObserved(): | ||
| klog.V(4).Info("FeatureGates initialized") | ||
| case <-time.After(1 * time.Minute): | ||
| klog.Error(errors.New("timed out waiting for FeatureGate detection"), "unable to start operator") | ||
| return | ||
| } | ||
|
|
||
| for i := 0; i < workers; i++ { | ||
| go wait.Until(optr.worker, time.Second, stopCh) | ||
| } | ||
|
|
@@ -353,19 +400,6 @@ func (optr *Operator) sync(key string) (reconcile.Result, error) { | |
| return optr.syncAll(operatorConfig) | ||
| } | ||
|
|
||
| func (optr *Operator) maoConfigFromInfrastructure() (*OperatorConfig, error) { | ||
| infra, err := optr.osClient.ConfigV1().Infrastructures().Get(context.Background(), "cluster", metav1.GetOptions{}) | ||
| if err != nil { | ||
|
|
@@ -382,12 +416,7 @@ func (optr *Operator) maoConfigFromInfrastructure() (*OperatorConfig, error) { | |
| return nil, err | ||
| } | ||
|
|
||
| providerControllerImage, err := getProviderControllerFromImages(provider, *images) | ||
| if err != nil { | ||
| return nil, err | ||
| } | ||
|
|
@@ -412,14 +441,25 @@ func (optr *Operator) maoConfigFromInfrastructure() (*OperatorConfig, error) { | |
| return nil, err | ||
| } | ||
|
|
||
| // in case the MHC controller is disabled, leave its image empty | ||
| mhcImage := machineAPIOperatorImage | ||
| featureGates, err := optr.featureGateAccessor.CurrentFeatureGates() | ||
| if err != nil { | ||
| return nil, err | ||
| } | ||
| if featureGates.Enabled(osconfigv1.FeatureGateMachineAPIOperatorDisableMachineHealthCheckController) { | ||
| klog.V(2).Info("Disabling MHC controller") | ||
| mhcImage = "" | ||
| } | ||
|
|
||
| return &OperatorConfig{ | ||
| TargetNamespace: optr.namespace, | ||
| Proxy: clusterWideProxy, | ||
| Controllers: Controllers{ | ||
| Provider: providerControllerImage, | ||
| MachineSet: machineAPIOperatorImage, | ||
| NodeLink: machineAPIOperatorImage, | ||
| MachineHealthCheck: mhcImage, | ||
| KubeRBACProxy: kubeRBACProxy, | ||
| TerminationHandler: terminationHandlerImage, | ||
| }, | ||
|
|
||
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
needed for the getting the controller ref for the recorder for the feature gate accessor...