support fix containerrecreaterequest

apis/apps/v1alpha1/containerrecreaterequest_types.go

@@ -106,6 +106,8 @@ type ContainerRecreateRequestStrategy struct {
 	FailurePolicy ContainerRecreateRequestFailurePolicyType `json:"failurePolicy,omitempty"`
 	// OrderedRecreate indicates whether to recreate the next container only if the previous one has recreated completely.
 	OrderedRecreate bool `json:"orderedRecreate,omitempty"`
+	// ForceRecreate indicates whether to force kill the container even if the previous container is starting.
+	ForceRecreate bool `json:"forceRecreate,omitempty"`
 	// TerminationGracePeriodSeconds is the optional duration in seconds to wait the container terminating gracefully.
 	// Value must be non-negative integer. The value zero indicates delete immediately.
 	// If this value is nil, we will use pod.Spec.TerminationGracePeriodSeconds as default value.
@@ -158,6 +160,8 @@ type ContainerRecreateRequestContainerRecreateState struct {
 	Phase ContainerRecreateRequestPhase `json:"phase"`
 	// A human readable message indicating details about this state.
 	Message string `json:"message,omitempty"`
+	// Container has been killed by force
+	IsForceKilled bool `json:"isForceKilled,omitempty"`
 }
 
 // ContainerRecreateRequestSyncContainerStatus only uses in the annotation `crr.apps.kruise.io/sync-container-statuses`.

config/crd/bases/apps.kruise.io_containerrecreaterequests.yaml

@@ -236,6 +236,10 @@ spec:
                     description: FailurePolicy decides whether to continue if one
                       container fails to recreate
                     type: string
+                  forceRecreate:
+                    description: ForceRecreate indicates whether to force kill the
+                      container even if the previous container is starting.
+                    type: boolean
                   minStartedSeconds:
                     description: Minimum number of seconds for which a newly created
                       container should be started and ready without any of its container
@@ -289,6 +293,9 @@ spec:
                   description: ContainerRecreateRequestContainerRecreateState contains
                     the recreation state of the container.
                   properties:
+                    isForceKilled:
+                      description: Container has been killed by force
+                      type: boolean
                     message:
                       description: A human readable message indicating details about
                         this state.

pkg/daemon/containerrecreate/crr_daemon_util.go

@@ -69,6 +69,8 @@ func getCurrentCRRContainersRecreateStates(
 	}
 
 	syncContainerStatuses := getCRRSyncContainerStatuses(crr)
+	// the statuses of force killed containers for this crr resource
+	forceKillContainerStatuses := getKilledContainersStatuses(crr)
 	var statuses []appsv1alpha1.ContainerRecreateRequestContainerRecreateState
 
 	for i := range crr.Spec.Containers {
@@ -84,6 +86,8 @@ func getCurrentCRRContainersRecreateStates(
 		syncContainerStatus := syncContainerStatuses[c.Name]
 		kubeContainerStatus := podStatus.FindContainerStatusByName(c.Name)
 
+		killedContainerStatus := forceKillContainerStatuses[c.Name]
+
 		var currentState appsv1alpha1.ContainerRecreateRequestContainerRecreateState
 		if kubeContainerStatus == nil {
 			// not found the real container
@@ -93,13 +97,21 @@ func getCurrentCRRContainersRecreateStates(
 				Message: "not found container on Node",
 			}
 
+		} else if crr.Spec.Strategy.ForceRecreate && containerIsNotForceKilled(killedContainerStatus) {
+			currentState = appsv1alpha1.ContainerRecreateRequestContainerRecreateState{
+				Name:          c.Name,
+				IsForceKilled: true,
+				Phase:         appsv1alpha1.ContainerRecreateRequestPending,
+			}
 		} else if kubeContainerStatus.State != kubeletcontainer.ContainerStateRunning {
 			// for no-running state, we consider it will be recreated or restarted soon
 			currentState = appsv1alpha1.ContainerRecreateRequestContainerRecreateState{
 				Name:  c.Name,
 				Phase: appsv1alpha1.ContainerRecreateRequestRecreating,
 			}
-
+			if crr.Spec.Strategy.ForceRecreate {
+				currentState.IsForceKilled = true
+			}
 		} else if kubeContainerStatus.ID.String() != c.StatusContext.ContainerID ||
 			kubeContainerStatus.RestartCount > int(c.StatusContext.RestartCount) ||
 			kubeContainerStatus.StartedAt.After(crr.CreationTimestamp.Time) {
@@ -114,12 +126,17 @@ func getCurrentCRRContainersRecreateStates(
 				syncContainerStatus.Ready {
 				currentState.Phase = appsv1alpha1.ContainerRecreateRequestSucceeded
 			}
-
+			if crr.Spec.Strategy.ForceRecreate {
+				currentState.IsForceKilled = true
+			}
 		} else {
 			currentState = appsv1alpha1.ContainerRecreateRequestContainerRecreateState{
 				Name:  c.Name,
 				Phase: appsv1alpha1.ContainerRecreateRequestPending,
 			}
+			if crr.Spec.Strategy.ForceRecreate {
+				currentState.IsForceKilled = true
+			}
 		}
 
 		statuses = append(statuses, currentState)
@@ -128,6 +145,16 @@ func getCurrentCRRContainersRecreateStates(
 	return statuses
 }
 
+func containerIsNotForceKilled(recreateState *appsv1alpha1.ContainerRecreateRequestContainerRecreateState) bool {
+	if recreateState == nil {
+		return true
+	}
+	if !recreateState.IsForceKilled {
+		return true
+	}
+	return false
+}
+
 func getCRRContainerRecreateState(crr *appsv1alpha1.ContainerRecreateRequest, name string) *appsv1alpha1.ContainerRecreateRequestContainerRecreateState {
 	for i := range crr.Status.ContainerRecreateStates {
 		c := &crr.Status.ContainerRecreateStates[i]
@@ -157,6 +184,19 @@ func getCRRSyncContainerStatuses(crr *appsv1alpha1.ContainerRecreateRequest) map
 	return statuses
 }
 
+func getKilledContainersStatuses(crr *appsv1alpha1.ContainerRecreateRequest) map[string]*appsv1alpha1.ContainerRecreateRequestContainerRecreateState {
+	crrContainerRecreateStatus := crr.Status.ContainerRecreateStates
+	forceKillStatuses := make(map[string]*appsv1alpha1.ContainerRecreateRequestContainerRecreateState, len(crrContainerRecreateStatus))
+	for i := range crrContainerRecreateStatus {
+		c := &crrContainerRecreateStatus[i]
+		if !c.IsForceKilled {
+			continue
+		}
+		forceKillStatuses[c.Name] = c
+	}
+	return forceKillStatuses
+}
+
 func convertCRRToPod(crr *appsv1alpha1.ContainerRecreateRequest) *v1.Pod {
 	podName := crr.Spec.PodName
 	podUID := types.UID(crr.Labels[appsv1alpha1.ContainerRecreateRequestPodUIDKey])