[Feat] Support StormService pause rollout in upgrade

* Update stormservice golang client
* Improve the test coverage
* Refactor the API to support manual resume
* improve the canary features
* Leave e2e test to future PRs
* fix lint and verify issues
* Polish the canary status

Signed-off-by: Jiaxin Shan <[email protected]>

Author: Jeffwan

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

@@ -17,6 +17,9 @@ limitations under the License.
 package v1alpha1
 
 import (
+	"strconv"
+	"time"
+
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	"k8s.io/apimachinery/pkg/util/intstr"
 )
@@ -119,6 +122,10 @@ type StormServiceStatus struct {
 
 	// The label selector information of the pods belonging to the StormService object.
 	ScalingTargetSelector string `json:"scalingTargetSelector,omitempty"`
+
+	// CanaryStatus tracks the progress of canary deployments.
+	// +optional
+	CanaryStatus *CanaryStatus `json:"canaryStatus,omitempty"`
 }
 
 // These are valid conditions of a stormService.
@@ -146,6 +153,10 @@ 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
@@ -185,6 +196,156 @@ 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
+	// +optional
+	Duration *intstr.IntOrString `json:"duration,omitempty"`
+}
+
+// DurationSeconds converts the pause duration to seconds
+// Returns:
+// - >= 0: pause duration in seconds
+// - 0: manual pause (nil duration) or resume (duration "0"/0)
+// - -1: invalid duration string
+func (p *PauseStep) DurationSeconds() int32 {
+	if p.Duration == nil {
+		return 0 // Manual pause
+	}
+
+	if p.Duration.Type == intstr.String {
+		// Try parsing as integer first
+		if s, err := strconv.ParseInt(p.Duration.StrVal, 10, 32); err == nil {
+			return int32(s)
+		}
+		// Try parsing as duration string
+		if d, err := time.ParseDuration(p.Duration.StrVal); err == nil {
+			return int32(d.Seconds())
+		}
+		return -1 // Invalid string
+	}
+
+	return p.Duration.IntVal
+}
+
+// IsManualPause returns true if this is a manual pause (nil duration)
+func (p *PauseStep) IsManualPause() bool {
+	return p.Duration == nil
+}
+
+// IsResume returns true if this represents a resume action (duration 0 or "0")
+func (p *PauseStep) IsResume() bool {
+	if p.Duration == nil {
+		return false
+	}
+	return p.DurationSeconds() == 0
+}
+
+// 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"`
+
+	// StableRevision is the revision of the stable/old version
+	// +optional
+	StableRevision string `json:"stableRevision,omitempty"`
+
+	// CanaryRevision is the revision of the canary/new version
+	// +optional
+	CanaryRevision string `json:"canaryRevision,omitempty"`
+
+	// Phase indicates the current phase of the canary deployment
+	// +optional
+	Phase CanaryPhase `json:"phase,omitempty"`
+
+	// CanaryReplicas is the number of RoleSets on canary version (replica mode)
+	// +optional
+	CanaryReplicas int32 `json:"canaryReplicas,omitempty"`
+
+	// StableReplicas is the number of RoleSets on stable version (replica mode)
+	// +optional
+	StableReplicas int32 `json:"stableReplicas,omitempty"`
+
+	// RoleCanaryCounts tracks per-role canary pod counts (pooled mode)
+	// +optional
+	RoleCanaryCounts map[string]int32 `json:"roleCanaryCounts,omitempty"`
+
+	// TotalCanaryPods is the total number of canary pods across all roles (pooled mode)
+	// +optional
+	TotalCanaryPods int32 `json:"totalCanaryPods,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{})
 }