[Feature] Support recreate pods for RayCluster using RayClusterSpec

Signed-off-by: win5923 <[email protected]>

Author: win5923

docs/reference/api.md

@@ -297,6 +297,7 @@ _Appears in:_
 
 | Field | Description | Default | Validation |
 | --- | --- | --- | --- |
+| `upgradeStrategy` _[RayClusterUpgradeStrategy](#rayclusterupgradestrategy)_ | UpgradeStrategy defines the scaling policy used when upgrading the RayCluster |  |  |
 | `authOptions` _[AuthOptions](#authoptions)_ | AuthOptions specifies the authentication options for the RayCluster. |  |  |
 | `suspend` _boolean_ | Suspend indicates whether a RayCluster should be suspended.<br />A suspended RayCluster will have head pods and worker pods deleted. |  |  |
 | `managedBy` _string_ | ManagedBy is an optional configuration for the controller or entity that manages a RayCluster.<br />The value must be either 'ray.io/kuberay-operator' or 'kueue.x-k8s.io/multikueue'.<br />The kuberay-operator reconciles a RayCluster which doesn't have this field at all or<br />the field value is the reserved string 'ray.io/kuberay-operator',<br />but delegates reconciling the RayCluster with 'kueue.x-k8s.io/multikueue' to the Kueue.<br />The field is immutable. |  |  |
@@ -309,6 +310,35 @@ _Appears in:_
 | `workerGroupSpecs` _[WorkerGroupSpec](#workergroupspec) array_ | WorkerGroupSpecs are the specs for the worker pods |  |  |
 
 
+#### RayClusterUpgradeStrategy
+
+
+
+
+
+
+
+_Appears in:_
+- [RayClusterSpec](#rayclusterspec)
+
+| Field | Description | Default | Validation |
+| --- | --- | --- | --- |
+| `type` _[RayClusterUpgradeType](#rayclusterupgradetype)_ | Type represents the strategy used when upgrading the RayCluster Pods. Currently supports `Recreate` and `None`. |  |  |
+
+
+#### RayClusterUpgradeType
+
+_Underlying type:_ _string_
+
+
+
+
+
+_Appears in:_
+- [RayClusterUpgradeStrategy](#rayclusterupgradestrategy)
+
+
+
 #### RayJob
 
 

helm-chart/kuberay-operator/crds/ray.io_rayclusters.yaml

@@ -11,6 +11,9 @@ import (
 
 // RayClusterSpec defines the desired state of RayCluster
 type RayClusterSpec struct {
+	// UpgradeStrategy defines the scaling policy used when upgrading the RayCluster
+	// +optional
+	UpgradeStrategy *RayClusterUpgradeStrategy `json:"upgradeStrategy,omitempty"`
 	// AuthOptions specifies the authentication options for the RayCluster.
 	// +optional
 	AuthOptions *AuthOptions `json:"authOptions,omitempty"`
@@ -49,6 +52,21 @@ type RayClusterSpec struct {
 	WorkerGroupSpecs []WorkerGroupSpec `json:"workerGroupSpecs,omitempty"`
 }
 
+type RayClusterUpgradeType string
+
+const (
+	// During upgrade, Recreate strategy will delete all existing pods before creating new ones
+	Recreate RayClusterUpgradeType = "Recreate"
+	// No new pod will be created while the strategy is set to None
+	RayClusterUpgradeNone RayClusterUpgradeType = "None"
+)
+
+type RayClusterUpgradeStrategy struct {
+	// Type represents the strategy used when upgrading the RayCluster Pods. Currently supports `Recreate` and `None`.
+	// +optional
+	Type *RayClusterUpgradeType `json:"type,omitempty"`
+}
+
 // AuthMode describes the authentication mode for the Ray cluster.
 type AuthMode string
 

helm-chart/kuberay-operator/crds/ray.io_rayjobs.yaml

@@ -28,7 +28,7 @@ const (
 	// During upgrade, NewCluster strategy will create new upgraded cluster and switch to it when it becomes ready
 	NewCluster RayServiceUpgradeType = "NewCluster"
 	// No new cluster will be created while the strategy is set to None
-	None RayServiceUpgradeType = "None"
+	RayServiceUpgradeNone RayServiceUpgradeType = "None"
 )
 
 // These statuses should match Ray Serve's application statuses