feat: change usage stats enablement to cluster config param

Author: kryanbeane

docs/sphinx/user-docs/cluster-configuration.rst

@@ -47,7 +47,7 @@ requirements for creating the Ray Cluster.
 Ray Usage Statistics
 -------------------
 
-By default, Ray usage statistics collection is disabled in CodeFlare SDK clusters. This stops statistics from being sent to AnyScale. If you want to enable usage statistics collection, you can set the ``RAY_USAGE_STATS_ENABLED`` environment variable to ``1`` in your cluster configuration:
+By default, Ray usage statistics collection is **disabled** in Ray Clusters created with the Codeflare SDK. This prevents statistics from being captured and sent externally. If you want to enable usage statistics collection, you can simply set the ``enable_usage_stats`` parameter to ``True`` in your cluster configuration:
 
 .. code:: python
 
@@ -56,9 +56,11 @@ By default, Ray usage statistics collection is disabled in CodeFlare SDK cluster
    cluster = Cluster(ClusterConfiguration(
        name='ray-example',
        namespace='default',
-       envs={'RAY_USAGE_STATS_ENABLED': '1'}  # Enable usage statistics
+       enable_usage_stats=True
    ))
 
+This will automatically set the ``RAY_USAGE_STATS_ENABLED`` environment variable to ``1`` for all Ray pods in the cluster. If you do not set this parameter, usage statistics will remain disabled (``RAY_USAGE_STATS_ENABLED=0``).
+
 The ``labels={"exampleLabel": "example"}`` parameter can be used to
 apply additional labels to the RayCluster resource.
 

src/codeflare_sdk/ray/cluster/config.py

@@ -102,6 +102,8 @@ class ClusterConfiguration:
             A list of V1VolumeMount objects to add to the Cluster
         enable_gcs_ft:
             A boolean indicating whether to enable GCS fault tolerance.
+        enable_usage_stats:
+            A boolean indicating whether to capture and send Ray usage stats externally.
         redis_address:
             The address of the Redis server to use for GCS fault tolerance, required when enable_gcs_ft is True.
         redis_password_secret:
@@ -151,6 +153,7 @@ class ClusterConfiguration:
     volumes: list[V1Volume] = field(default_factory=list)
     volume_mounts: list[V1VolumeMount] = field(default_factory=list)
     enable_gcs_ft: bool = False
+    enable_usage_stats: bool = False
     redis_address: Optional[str] = None
     redis_password_secret: Optional[Dict[str, str]] = None
     external_storage_namespace: Optional[str] = None
@@ -161,8 +164,9 @@ def __post_init__(self):
                 "Warning: TLS verification has been disabled - Endpoint checks will be bypassed"
             )
 
-        # Set default environment variable to disable Ray usage stats if not already set
-        if "RAY_USAGE_STATS_ENABLED" not in self.envs:
+        if self.enable_usage_stats:
+            self.envs["RAY_USAGE_STATS_ENABLED"] = "1"
+        else:
             self.envs["RAY_USAGE_STATS_ENABLED"] = "0"
 
         if self.enable_gcs_ft:

src/codeflare_sdk/ray/cluster/test_config.py

@@ -238,7 +238,7 @@ def test_ray_usage_stats_enabled(mocker):
         ClusterConfiguration(
             name="usage-stats-enabled-cluster",
             namespace="ns",
-            envs={"RAY_USAGE_STATS_ENABLED": "1"},
+            enable_usage_stats=True,
         )
     )