ray-project
diff --git a/‎docs/reference/api.md‎
Lines changed: 20 additions & 0 deletions b/‎docs/reference/api.md‎
Lines changed: 20 additions & 0 deletions
diff --git a/‎go.mod‎
Lines changed: 6 additions & 5 deletions b/‎go.mod‎
Lines changed: 6 additions & 5 deletions
diff --git a/‎go.sum‎
Lines changed: 12 additions & 9 deletions b/‎go.sum‎
Lines changed: 12 additions & 9 deletions
diff --git a/‎helm-chart/kuberay-operator/README.md‎
Lines changed: 2 additions & 0 deletions b/‎helm-chart/kuberay-operator/README.md‎
Lines changed: 2 additions & 0 deletions
diff --git a/‎helm-chart/kuberay-operator/crds/ray.io_rayservices.yaml‎
Lines changed: 37 additions & 0 deletions b/‎helm-chart/kuberay-operator/crds/ray.io_rayservices.yaml‎
Lines changed: 37 additions & 0 deletions
diff --git a/‎helm-chart/kuberay-operator/templates/_helpers.tpl‎
Lines changed: 11 additions & 0 deletions b/‎helm-chart/kuberay-operator/templates/_helpers.tpl‎
Lines changed: 11 additions & 0 deletions
diff --git a/‎helm-chart/kuberay-operator/values.yaml‎
Lines changed: 2 additions & 0 deletions b/‎helm-chart/kuberay-operator/values.yaml‎
Lines changed: 2 additions & 0 deletions
diff --git a/‎ray-operator/Makefile‎
Lines changed: 9 additions & 1 deletion b/‎ray-operator/Makefile‎
Lines changed: 9 additions & 1 deletion
diff --git a/‎ray-operator/apis/ray/v1/rayservice_types.go‎
Lines changed: 35 additions & 2 deletions b/‎ray-operator/apis/ray/v1/rayservice_types.go‎
Lines changed: 35 additions & 2 deletions
diff --git a/‎ray-operator/apis/ray/v1/zz_generated.deepcopy.go‎
Lines changed: 49 additions & 0 deletions b/‎ray-operator/apis/ray/v1/zz_generated.deepcopy.go‎
Lines changed: 49 additions & 0 deletions
@@ -55,6 +55,25 @@ _Appears in:_
 
 
 
+#### ClusterUpgradeOptions
+
+
+
+These options are currently only supported for the IncrementalUpgrade type.
+
+
+
+_Appears in:_
+- [RayServiceUpgradeStrategy](#rayserviceupgradestrategy)
+
+| Field | Description | Default | Validation |
+| --- | --- | --- | --- |
+| `maxSurgePercent` _integer_ | The capacity of serve requests the upgraded cluster should scale to handle each interval.<br />Defaults to 100%. | 100 |  |
+| `stepSizePercent` _integer_ | The percentage of traffic to switch to the upgraded RayCluster at a set interval after scaling by MaxSurgePercent. |  |  |
+| `intervalSeconds` _integer_ | The interval in seconds between transferring StepSize traffic from the old to new RayCluster. |  |  |
+| `gatewayClassName` _string_ | The name of the Gateway Class installed by the Kubernetes Cluster admin. |  |  |
+
+
 #### DeletionCondition
 
 
@@ -377,6 +396,7 @@ _Appears in:_
 | Field | Description | Default | Validation |
 | --- | --- | --- | --- |
 | `type` _[RayServiceUpgradeType](#rayserviceupgradetype)_ | Type represents the strategy used when upgrading the RayService. Currently supports `NewCluster` and `None`. |  |  |
+| `clusterUpgradeOptions` _[ClusterUpgradeOptions](#clusterupgradeoptions)_ | ClusterUpgradeOptions defines the behavior of a NewClusterWithIncrementalUpgrade type.<br />RayServiceIncrementalUpgrade feature gate must be enabled to set ClusterUpgradeOptions. |  |  |
 
 
 #### RayServiceUpgradeType
 
@@ -73,7 +73,7 @@ require (
 	github.com/liggitt/tabwriter v0.0.0-20181228230101-89fcab3d43de // indirect
 	github.com/mailru/easyjson v0.9.0 // indirect
 	github.com/mattn/go-colorable v0.1.13 // indirect
-	github.com/mattn/go-isatty v0.0.19 // indirect
+	github.com/mattn/go-isatty v0.0.20 // indirect
 	github.com/mitchellh/go-wordwrap v1.0.1 // indirect
 	github.com/moby/spdystream v0.5.0 // indirect
 	github.com/moby/term v0.5.0 // indirect
@@ -95,12 +95,12 @@ require (
 	go.uber.org/automaxprocs v1.6.0 // indirect
 	go.uber.org/multierr v1.11.0 // indirect
 	go.uber.org/zap v1.27.0 // indirect
-	golang.org/x/net v0.38.0 // indirect
+	golang.org/x/net v0.39.0 // indirect
 	golang.org/x/oauth2 v0.27.0 // indirect
-	golang.org/x/sync v0.12.0 // indirect
+	golang.org/x/sync v0.13.0 // indirect
 	golang.org/x/sys v0.32.0 // indirect
-	golang.org/x/term v0.30.0 // indirect
-	golang.org/x/text v0.23.0 // indirect
+	golang.org/x/term v0.31.0 // indirect
+	golang.org/x/text v0.24.0 // indirect
 	golang.org/x/time v0.10.0 // indirect
 	golang.org/x/tools v0.31.0 // indirect
 	gomodules.xyz/jsonpatch/v2 v2.4.0 // indirect
@@ -112,6 +112,7 @@ require (
 	k8s.io/component-base v0.33.1 // indirect
 	k8s.io/component-helpers v0.33.1 // indirect
 	k8s.io/kube-openapi v0.0.0-20250318190949-c8a335a9a2ff // indirect
+	sigs.k8s.io/gateway-api v1.3.0 // indirect
 	sigs.k8s.io/json v0.0.0-20241014173422-cfa47c3a1cc8 // indirect
 	sigs.k8s.io/kustomize/api v0.19.0 // indirect
 	sigs.k8s.io/kustomize/kyaml v0.19.0 // indirect
 
@@ -174,6 +174,8 @@ spec:
 | featureGates[1].enabled | bool | `false` |  |
 | featureGates[2].name | string | `"RayMultiHostIndexing"` |  |
 | featureGates[2].enabled | bool | `false` |  |
+| featureGates[3].name | string | `"RayServiceIncrementalUpgrade"` |  |
+| featureGates[3].enabled | bool | `false` |  |
 | metrics.enabled | bool | `true` | Whether KubeRay operator should emit control plane metrics. |
 | metrics.serviceMonitor.enabled | bool | `false` | Enable a prometheus ServiceMonitor |
 | metrics.serviceMonitor.interval | string | `"30s"` | Prometheus ServiceMonitor interval |
 
@@ -222,6 +222,17 @@ rules:
   - patch
   - update
   - watch
+- apiGroups:
+  - gateway.networking.k8s.io
+  resources:
+  - gateways
+  - httproutes
+  verbs:
+  - create
+  - get
+  - list
+  - update
+  - watch
 - apiGroups:
   - networking.k8s.io
   resources:
 
@@ -119,6 +119,8 @@ featureGates:
   enabled: false
 - name: RayMultiHostIndexing
   enabled: false
+- name: RayServiceIncrementalUpgrade
+  enabled: false
 
 # Configurations for KubeRay operator metrics.
 metrics:
 
@@ -75,8 +75,16 @@ test-e2e-autoscaler: WHAT ?= ./test/e2eautoscaler
 test-e2e-autoscaler: manifests fmt vet ## Run e2e autoscaler tests.
 	go test -timeout 30m -v $(WHAT)
 
+test-e2e-rayservice: WHAT ?= ./test/e2erayservice
+test-e2e-rayservice: manifests fmt vet ## Run e2e RayService tests.
+	go test -timeout 30m -v $(WHAT)
+
 test-e2e-upgrade: WHAT ?= ./test/e2eupgrade
-test-e2e-upgrade: manifests fmt vet ## Run e2e tests.
+test-e2e-upgrade: manifests fmt vet ## Run e2e operator upgrade tests.
+	go test -timeout 30m -v $(WHAT)
+
+test-e2e-incremental-upgrade: WHAT ?= ./test/e2eincrementalupgrade
+test-e2e-incremental-upgrade: manifests fmt vet ## Run e2e RayService incremental upgrade tests.
 	go test -timeout 30m -v $(WHAT)
 
 test-e2e-rayjob-submitter: WHAT ?= ./test/e2erayjobsubmitter
 
@@ -22,6 +22,9 @@ const (
 type RayServiceUpgradeType string
 
 const (
+	// During upgrade, NewClusterWithIncrementalUpgrade strategy will create an upgraded cluster to gradually scale
+	// and migrate traffic to using Gateway API.
+	NewClusterWithIncrementalUpgrade RayServiceUpgradeType = "NewClusterWithIncrementalUpgrade"
 	// 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
@@ -57,10 +60,27 @@ var DeploymentStatusEnum = struct {
 	UNHEALTHY: "UNHEALTHY",
 }
 
+// These options are currently only supported for the IncrementalUpgrade type.
+type ClusterUpgradeOptions struct {
+	// The capacity of serve requests the upgraded cluster should scale to handle each interval.
+	// Defaults to 100%.
+	// +kubebuilder:default:=100
+	MaxSurgePercent *int32 `json:"maxSurgePercent,omitempty"`
+	// The percentage of traffic to switch to the upgraded RayCluster at a set interval after scaling by MaxSurgePercent.
+	StepSizePercent *int32 `json:"stepSizePercent"`
+	// The interval in seconds between transferring StepSize traffic from the old to new RayCluster.
+	IntervalSeconds *int32 `json:"intervalSeconds"`
+	// The name of the Gateway Class installed by the Kubernetes Cluster admin.
+	GatewayClassName string `json:"gatewayClassName"`
+}
+
 type RayServiceUpgradeStrategy struct {
 	// Type represents the strategy used when upgrading the RayService. Currently supports `NewCluster` and `None`.
 	// +optional
 	Type *RayServiceUpgradeType `json:"type,omitempty"`
+	// ClusterUpgradeOptions defines the behavior of a NewClusterWithIncrementalUpgrade type.
+	// RayServiceIncrementalUpgrade feature gate must be enabled to set ClusterUpgradeOptions.
+	ClusterUpgradeOptions *ClusterUpgradeOptions `json:"clusterUpgradeOptions,omitempty"`
 }
 
 // RayServiceSpec defines the desired state of RayService
@@ -129,6 +149,20 @@ type RayServiceStatus struct {
 	// Important: Run "make" to regenerate code after modifying this file
 	// +optional
 	Applications map[string]AppStatus `json:"applicationStatuses,omitempty"`
+	// TargetCapacity is the `target_capacity` percentage for all Serve replicas
+	// across the cluster for this RayService. The `num_replicas`, `min_replicas`, `max_replicas`,
+	// and `initial_replicas` for each deployment will be scaled by this percentage."
+	// +optional
+	TargetCapacity *int32 `json:"targetCapacity,omitempty"`
+	// TrafficRoutedPercent is the percentage of traffic that is routed to the Serve service
+	// for this RayService. TrafficRoutedPercent is updated to reflect the weight on the HTTPRoute
+	// created for this RayService during incremental upgrades to a new cluster.
+	// +optional
+	TrafficRoutedPercent *int32 `json:"trafficRoutedPercent,omitempty"`
+	// LastTrafficMigratedTime is the last time that TrafficRoutedPercent was updated to a new value
+	// for this RayService.
+	// +optional
+	LastTrafficMigratedTime *metav1.Time `json:"lastTrafficMigratedTime,omitempty"`
 	// +optional
 	RayClusterName string `json:"rayClusterName,omitempty"`
 	// +optional
@@ -184,8 +218,7 @@ const (
 type RayService struct {
 	metav1.TypeMeta   `json:",inline"`
 	metav1.ObjectMeta `json:"metadata,omitempty"`
-
-	Spec RayServiceSpec `json:"spec,omitempty"`
+	Spec              RayServiceSpec `json:"spec,omitempty"`
 	// +optional
 	Status RayServiceStatuses `json:"status,omitempty"`
 }