WIP: [Feat] Support StormService pause rollout in upgrade

Makefile

@@ -14,7 +14,7 @@ AIBRIX_IMAGES := $(foreach img,$(IMAGES),$(AIBRIX_CONTAINER_REGISTRY_NAMESPACE)/
 IMG ?= ${AIBRIX_CONTAINER_REGISTRY_NAMESPACE}/controller-manager:${IMAGE_TAG}
 
 # ENVTEST_K8S_VERSION refers to the version of kubebuilder assets to be downloaded by envtest binary.
-ENVTEST_K8S_VERSION = 1.29.0
+ENVTEST_K8S_VERSION = 1.30.0
 
 # Get the currently used golang install path (in GOPATH/bin, unless GOBIN is set)
 ifeq (,$(shell go env GOBIN))

api/orchestration/v1alpha1/stormservice_types.go

@@ -131,6 +131,10 @@ type StormServiceStatus struct {
 	//
 	// +optional
 	RoleStatuses []RoleStatus `json:"roleStatuses,omitempty"`
+
+	// CanaryStatus tracks the progress of canary deployments.
+	// +optional
+	CanaryStatus *CanaryStatus `json:"canaryStatus,omitempty"`
 }
 
 // These are valid conditions of a stormService.
@@ -158,6 +162,10 @@ type StormServiceUpdateStrategy struct {
 
 	// +optional
 	MaxSurge *intstr.IntOrString `json:"maxSurge,omitempty" protobuf:"bytes,2,opt,name=maxSurge"`
+
+	// Canary defines the canary deployment strategy for gradual rollouts.
+	// +optional
+	Canary *CanaryUpdateStrategy `json:"canary,omitempty"`
 }
 
 // +enum
@@ -197,6 +205,105 @@ type StormServiceList struct {
 	Items           []StormService `json:"items"`
 }
 
+// CanaryUpdateStrategy defines the canary deployment configuration
+type CanaryUpdateStrategy struct {
+	// Steps defines the sequence of canary deployment steps
+	Steps []CanaryStep `json:"steps,omitempty"`
+}
+
+// CanaryStep defines a single step in the canary deployment process
+type CanaryStep struct {
+	// SetWeight defines the percentage of traffic/replicas to route to the new version
+	// +kubebuilder:validation:Minimum=0
+	// +kubebuilder:validation:Maximum=100
+	// +optional
+	SetWeight *int32 `json:"setWeight,omitempty"`
+
+	// Pause defines a pause in the canary deployment
+	// +optional
+	Pause *PauseStep `json:"pause,omitempty"`
+}
+
+// PauseStep defines pause behavior in canary deployments
+type PauseStep struct {
+	// Duration specifies how long to pause
+	// - String: "30s", "5m", etc. (parsed as time.Duration)
+	// - Int: seconds as integer
+	// - nil: manual pause requiring user intervention
+	// Resume manual pause by setting duration to "0" or 0
+
+	// Duration field is accepted but not implemented.
+	// At this moment, all pauses are manual and require removing the pause condition to resume.
+	// - pause: {} # this is accepted
+	// - pause:    # api accepted but not implemented.
+	//     duration: "60s"
+
+	// +optional
+	Duration *intstr.IntOrString `json:"duration,omitempty"`
+}
+
+// CanaryStatus tracks the progress of a canary deployment
+type CanaryStatus struct {
+	// CurrentStep is the index of the current step in the canary deployment
+	// +optional
+	CurrentStep int32 `json:"currentStep,omitempty"`
+
+	// PauseConditions indicates the reasons why the canary deployment is paused
+	// When paused, the first pause condition's StartTime indicates when the pause began
+	// +optional
+	PauseConditions []PauseCondition `json:"pauseConditions,omitempty"`
+
+	// NOTE: Removed StableRevision and CanaryRevision fields
+	// Use status.CurrentRevision for stable revision
+	// Use status.UpdateRevision for canary revision
+
+	// Phase indicates the current phase of the canary deployment
+	// +optional
+	Phase CanaryPhase `json:"phase,omitempty"`
+
+	// AbortedAt indicates when the canary deployment was aborted
+	// +optional
+	AbortedAt *metav1.Time `json:"abortedAt,omitempty"`
+
+	// Message provides details about the current canary state
+	// +optional
+	Message string `json:"message,omitempty"`
+}
+
+// CanaryPhase represents the phase of a canary deployment
+// +enum
+type CanaryPhase string
+
+const (
+	// CanaryPhaseInitializing indicates the canary deployment is starting
+	CanaryPhaseInitializing CanaryPhase = "Initializing"
+	// CanaryPhaseProgressing indicates the canary deployment is progressing through steps
+	CanaryPhaseProgressing CanaryPhase = "Progressing"
+	// CanaryPhasePaused indicates the canary deployment is paused
+	CanaryPhasePaused CanaryPhase = "Paused"
+	// CanaryPhaseCompleted indicates the canary deployment has completed successfully
+	CanaryPhaseCompleted CanaryPhase = "Completed"
+	// CanaryPhaseAborted indicates the canary deployment was aborted/rolled back
+	CanaryPhaseAborted CanaryPhase = "Aborted"
+)
+
+// PauseReason represents the reason for a pause condition
+// +enum
+type PauseReason string
+
+const (
+	// PauseReasonCanaryPauseStep indicates a pause at a canary step
+	PauseReasonCanaryPauseStep PauseReason = "CanaryPauseStep"
+)
+
+// PauseCondition represents a pause condition in the canary deployment
+type PauseCondition struct {
+	// Reason indicates why the canary deployment was paused
+	Reason PauseReason `json:"reason"`
+	// StartTime is when the pause condition was added
+	StartTime metav1.Time `json:"startTime"`
+}
+
 func init() {
 	SchemeBuilder.Register(&StormService{}, &StormServiceList{})
 }

config/crd/orchestration/orchestration.aibrix.ai_stormservices.yaml

@@ -4106,6 +4106,27 @@ spec:
                 type: object
               updateStrategy:
                 properties:
+                  canary:
+                    properties:
+                      steps:
+                        items:
+                          properties:
+                            pause:
+                              properties:
+                                duration:
+                                  anyOf:
+                                  - type: integer
+                                  - type: string
+                                  x-kubernetes-int-or-string: true
+                              type: object
+                            setWeight:
+                              format: int32
+                              maximum: 100
+                              minimum: 0
+                              type: integer
+                          type: object
+                        type: array
+                    type: object
                   maxSurge:
                     anyOf:
                     - type: integer
@@ -4129,6 +4150,32 @@ spec:
             type: object
           status:
             properties:
+              canaryStatus:
+                properties:
+                  abortedAt:
+                    format: date-time
+                    type: string
+                  currentStep:
+                    format: int32
+                    type: integer
+                  message:
+                    type: string
+                  pauseConditions:
+                    items:
+                      properties:
+                        reason:
+                          type: string
+                        startTime:
+                          format: date-time
+                          type: string
+                      required:
+                      - reason
+                      - startTime
+                      type: object
+                    type: array
+                  phase:
+                    type: string
+                type: object
               collisionCount:
                 format: int32
                 type: integer