[Helm] Add gcsFaultToleranceOptions in RayCluster chart (#3881)

* [Helm] Add gcsFaultToleranceOptions in RayCluster chart

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

* Reslove comments

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

---------

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

Author: win5923

helm-chart/ray-cluster/templates/raycluster-cluster.yaml

@@ -20,6 +20,21 @@ spec:
   autoscalerOptions:
     {{- toYaml . | nindent 4 }}
   {{- end }}
+  {{- if .Values.gcsFaultTolerance.enabled }}
+  gcsFaultToleranceOptions:
+    redisAddress: {{ .Values.gcsFaultTolerance.redisAddress | quote }}
+    {{- with .Values.gcsFaultTolerance.externalStorageNamespace }}
+    externalStorageNamespace: {{ . | quote }}
+    {{- end }}
+    {{- with .Values.gcsFaultTolerance.redisPassword }}
+    redisPassword:
+      {{- toYaml . | nindent 6 }}
+    {{- end }}
+    {{- with .Values.gcsFaultTolerance.redisUsername }}
+    redisUsername:
+      {{- toYaml . | nindent 6 }}
+    {{- end }}
+  {{- end }}
   headGroupSpec:
     {{- with .Values.head.headService }}
     headService:

helm-chart/ray-cluster/tests/raycluster_test.yaml

@@ -97,6 +97,33 @@ tests:
               seccompProfile:
                 type: RuntimeDefault
 
+  - it: Should add GCS fault tolerance options when `gcsFaultTolerance.enabled` is `true`
+    set:
+      gcsFaultTolerance:
+        enabled: true
+        redisAddress: "test-redis:6379"
+        externalStorageNamespace: "test-gcs-namespace"
+        redisPassword:
+          valueFrom:
+            secretKeyRef:
+              name: redis-secret
+              key: password
+        redisUsername:
+          value: "test-user"
+    asserts:
+      - equal:
+          path: spec.gcsFaultToleranceOptions
+          value:
+            redisAddress: "test-redis:6379"
+            externalStorageNamespace: "test-gcs-namespace"
+            redisPassword:
+              valueFrom:
+                secretKeyRef:
+                  name: redis-secret
+                  key: password
+            redisUsername:
+              value: "test-user"
+
   # ===========================================================================
   # Test Head Group
   # ===========================================================================

helm-chart/ray-cluster/values.yaml

@@ -16,6 +16,32 @@ fullnameOverride: ""
 imagePullSecrets: []
   # - name: an-existing-secret
 
+# gcsFaultTolerance specifies configuration for GCS fault tolerance.
+# See https://docs.ray.io/en/latest/cluster/kubernetes/user-guides/kuberay-gcs-ft.html#kuberay-gcs-ft
+# for more details.
+gcsFaultTolerance:
+  enabled: false
+  # In most cases, you don't need to set `externalStorageNamespace` because KubeRay will
+  # automatically set it to the UID of RayCluster. Only modify this annotation if you fully understand
+  # the behaviors of the Ray GCS FT and RayService to avoid misconfiguration.
+  # externalStorageNamespace: "my-raycluster-storage"
+  # redisAddress: "redis:6379"
+  # redisPassword:
+    # Reference to a secret key containing the Redis password
+    # Secret must be manually created in the namespace
+    # valueFrom:
+    #   secretKeyRef:
+    #       name: redis-password-secret
+    #       key: password
+  # redisUsername:
+    # Reference to a secret key containing the Redis username
+    # Secret must be manually created in the namespace
+    # valueFrom:
+    #   secretKeyRef:
+    #     name: redis-username-secret
+    #     key: username
+
+
 # common defined values shared between the head and worker
 common:
   # containerEnv specifies environment variables for the Ray head and worker containers.