Address PR feedback

Author: 02strich

openapi/openapiv2.json

@@ -13,25 +13,33 @@ import "temporal/api/compute/v1/provider.proto";
 import "temporal/api/compute/v1/scaler.proto";
 import "temporal/api/enums/v1/task_queue.proto";
 
+message ComputeConfigScalingGroup {
+    // Optional. The set of task queue types this scaling group serves. 
+    // If not provided, this scaling group serves all not otherwise defined
+    // task types.
+    repeated temporal.api.enums.v1.TaskQueueType task_queue_types = 1;
+
+    // Stores instructions for a worker control plane controller how to respond
+    // to worker lifeycle events.
+    temporal.api.compute.v1.ComputeProvider provider = 3;
+
+    // Informs a worker lifecycle controller *when* and *how often* to perform
+    // certain worker lifecycle actions like starting a serverless worker.
+    temporal.api.compute.v1.ComputeScaler scaler = 4;
+}
+
 // ComputeConfig stores configuration that helps a worker control plane
 // controller understand *when* and *how* to respond to worker lifecycle
 // events.
 message ComputeConfig {
-    message ComputeConfigEntry {
-        // The set of task queue types this compute config serves.
-        repeated temporal.api.enums.v1.TaskQueueType task_queue_types = 1;
-
-        // Stores instructions for a worker control plane controller how to respond
-        // to worker lifeycle events.
-        temporal.api.compute.v1.ComputeProvider provider = 2;
-
-        // Informs a worker lifecycle controller *when* and *how often* to perform
-        // certain worker lifecycle actions like starting a serverless worker.
-        temporal.api.compute.v1.ComputeScaler scaler = 3;
-    }
-
-    // Each entry describes a compute config for a specific subset of the deployment's scope.
-    // Entries might be seperated by different task queue types.
-    repeated ComputeConfigEntry entries = 1;
+
+    // Each scaling group describes a compute config for a specific subset of the worker 
+    // deployment version: covering a specific set of task types and/or regions.
+    // Having different configurations for different task types, allows independent
+    // tuning of activity and workflow task processing (for example).
+    //
+    // The key of the map is the ID of the scaling group used to reference it in subsequent
+    // update calls.
+    map<string, ComputeConfigScalingGroup> scaling_groups = 1;
 }
 

openapi/openapiv3.yaml

@@ -21,19 +21,14 @@ message ComputeProvider {
     // can be used by implementations to understand how to interpret the
     // contents of the provider_details field.
     string type = 1;
+
     // Contains provider-specific instructions and configuration.
-    oneof detail {
-        // will be an unencrypted, JSON-encoded object of provider-specific
-        // information
-        // (-- api-linter: core::0146::any=disabled
-        //     aip.dev/not-precedent: This needs to be extensible to
-        //     externally-written compute providers --)
-        string detail_json = 2;
-        // will be an encrypted, encoded bytestring containing
-        // provider-specific information. The implementation must understand
-        // how to decrypt the payload.
-        temporal.api.common.v1.Payload detail_payload = 3;
-    }
+    // For server-implemented providers, use the SDK's default content
+    // converter to ensure the server can understand it.
+    // For remote-implemented providers, you might use your own content
+    // converters according to what the remote endpoints understand.
+    temporal.api.common.v1.Payload details = 2;
+    
     // Optional. If the compute provider is a Nexus service, this should point
     // there.
     string nexus_endpoint = 10;

temporal/api/compute/v1/config.proto

@@ -14,23 +14,15 @@ import "temporal/api/common/v1/message.proto";
 // ComputeScaler instructs the Temporal Service when to scale up or down the number of
 // Workers that comprise a WorkerDeployment.
 message ComputeScaler {
-    // type of the compute scaler. this string is implementation-specific and
+    // Type of the compute scaler. this string is implementation-specific and
     // can be used by implementations to understand how to interpret the
     // contents of the scaler_details field.
     string type = 1;
 
     // Contains scaler-specific instructions and configuration.
-    oneof detail {
-        // will be an unencrypted, JSON-encoded object of scaler-specific
-        // information
-        // (-- api-linter: core::0146::any=disabled
-        //     aip.dev/not-precedent: This needs to be extensible to
-        //     externally-written compute scalers --)
-        string detail_json = 2;
-
-        // will be an encrypted, encoded bytestring containing
-        // scaler-specific information. The implementation must understand
-        // how to decrypt the payload.
-        temporal.api.common.v1.Payload detail_payload = 3;
-    }
+    // For server-implemented scalers, use the SDK's default content
+    // converter to ensure the server can understand it.
+    // For remote-implemented scalers, you might use your own content
+    // converters according to what the remote endpoints understand.
+    temporal.api.common.v1.Payload details = 2;
 }

temporal/api/compute/v1/provider.proto

@@ -216,10 +216,6 @@ message WorkerDeploymentInfo {
     // relevant task queues and their partitions.
     temporal.api.enums.v1.RoutingConfigUpdateState routing_config_update_state = 7;
 
-    // Contains information used by worker control plane controllers to handle
-    // scale events.
-    temporal.api.compute.v1.ComputeConfig compute_config = 20;
-
     message WorkerDeploymentVersionSummary {
         // Deprecated. Use `deployment_version`.
         string version = 1 [deprecated = true];

temporal/api/compute/v1/scaler.proto

@@ -1268,10 +1268,10 @@ message GetSystemInfoResponse {
         // This flag is dependent both on server version and for Nexus to be enabled via server configuration.
         bool nexus = 11;
 
-        // True if the server supports serverless deployments.
-        // This flag is dependent both on server version and for serverless deployments
+        // True if the server supports server-scaled deployments.
+        // This flag is dependent both on server version and for server-scaled deployments
         // to be enabled via server configuration.
-        bool serverless_deployments = 12;
+        bool server_scaled_deployments = 12;
 
     }
 }
@@ -2353,10 +2353,6 @@ message CreateWorkerDeploymentRequest {
     // this name already exists, an error will be returned.
     string deployment_name = 2;
 
-    // Optional. Contains the new worker compute configuration for the Worker
-    // Deployment. Used for worker scale management.
-    temporal.api.compute.v1.ComputeConfig compute_config = 3;
-
     // Optional. The identity of the client who initiated this request.
     string identity = 4;
     // A unique identifier for this create request for idempotence. Typically UUIDv4.
@@ -2402,20 +2398,16 @@ message CreateWorkerDeploymentVersionRequest {
     // Required.
     temporal.api.deployment.v1.WorkerDeploymentVersion deployment_version = 2;
 
-    // All the Task Queues that will be polled from this Deployment version.
-    repeated VersionTaskQueueInfo task_queue_infos = 6;
-    message VersionTaskQueueInfo {
-        string name = 1;
-        temporal.api.enums.v1.TaskQueueType type = 2;
-    }
-
     // Optional. Contains the new worker compute configuration for the Worker
     // Deployment. Used for worker scale management.
-    temporal.api.compute.v1.ComputeConfig compute_config = 3;
+    temporal.api.compute.v1.ComputeConfig compute_config = 4;
 
     // Optional. The identity of the client who initiated this request.
-    string identity = 4;
+    string identity = 3;
+    
     // A unique identifier for this create request for idempotence. Typically UUIDv4.
+    // If a second request with the same ID is recieved, it is considered a successful no-op.
+    // Retrying with a different request ID for the same deployment name + build ID is an error.
     string request_id = 5;
 }
 
@@ -2460,22 +2452,63 @@ message DeleteWorkerDeploymentRequest {
 message DeleteWorkerDeploymentResponse {
 }
 
-// Used to update the compute provider of a Worker Deployment Version.
-message UpdateWorkerDeploymentVersionComputeProviderRequest {
+// Used to update the compute config of a Worker Deployment Version.
+message UpdateWorkerDeploymentVersionComputeConfigRequest {
     string namespace = 1;
 
     // Required.
     temporal.api.deployment.v1.WorkerDeploymentVersion deployment_version = 2;
 
-    // Optional. Contains the new worker compute configuration for the Worker
-    // Deployment. Used for worker scale management.
-    temporal.api.compute.v1.ComputeConfig compute_config = 3;
+    message ScalingGroupUpdate {
+      temporal.api.compute.v1.ComputeConfigScalingGroup scaling_group = 1;
+
+      // Controls which fields from `scaling_group` will be applied
+      google.protobuf.FieldMask update_mask = 2;
+    }
+
+    // Optional. Contains the compute config scaling groups to add or updated for the Worker
+    // Deployment.
+    map<string, ScalingGroupUpdate> compute_config_scaling_groups = 6;
+
+    // Optional. Contains the compute config scaling groups to remove from the Worker Deployment.
+    repeated string remove_compute_config_scaling_groups = 7;
 
     // Optional. The identity of the client who initiated this request.
-    string identity = 4;
+    string identity = 3;
+
+    // A unique identifier for this create request for idempotence. Typically UUIDv4.
+    // If a second request with the same ID is recieved, it is considered a successful no-op.
+    // Retrying with a different request ID for the same deployment name + build ID is an error.
+    string request_id = 4;
+
+    // Optional. This value is returned so that it can be optionally passed to APIs
+    // that write to the Worker Deployment state to ensure that the state
+    // did not change between this API call and a future write.
+    bytes conflict_token = 5;
+}
+
+message UpdateWorkerDeploymentVersionComputeConfigResponse {
+    // This value is returned so that it can be optionally passed to APIs that
+    // write to the WorkerDeployment state to ensure that the state did not
+    // change between this API call and a future write.
+    bytes conflict_token = 1;
+}
+
+// Used to validate the compute config without attaching it to a Worker Deployment Version.
+message ValidateWorkerDeploymentVersionComputeConfigRequest {
+    string namespace = 1;
+
+    // Required.
+    temporal.api.deployment.v1.WorkerDeploymentVersion deployment_version = 2;
+
+    // Required. Contains the new worker compute configuration for the Worker Deployment.
+    temporal.api.compute.v1.ComputeConfig compute_config = 4;
+
+    // Optional. The identity of the client who initiated this request.
+    string identity = 3;
 }
 
-message UpdateWorkerDEploymentVersionComputeProviderResponse {
+message ValidateWorkerDeploymentVersionComputeConfigResponse {
 }
 
 // Used to update the user-defined metadata of a Worker Deployment Version.

temporal/api/deployment/v1/message.proto

@@ -1146,14 +1146,27 @@ service WorkflowService {
         };
     }
 
-    // Updates the compute provider attached to a Worker Deployment Version.
+    // Updates the compute config attached to a Worker Deployment Version.
     // Experimental. This API might significantly change or be removed in a future release.
-    rpc UpdateWorkerDeploymentVersionComputeProvider (UpdateWorkerDeploymentVersionComputeProviderRequest) returns (UpdateWorkerDEploymentVersionComputeProviderResponse) {
+    rpc UpdateWorkerDeploymentVersionComputeConfig (UpdateWorkerDeploymentVersionComputeConfigRequest) returns (UpdateWorkerDeploymentVersionComputeConfigResponse) {
         option (google.api.http) = {
-            post: "/namespaces/{namespace}/worker-deployment-versions/{deployment_version.deployment_name}/{deployment_version.build_id}/update-compute"
+            post: "/namespaces/{namespace}/worker-deployment-versions/{deployment_version.deployment_name}/{deployment_version.build_id}/update-compute-config"
             body: "*"
             additional_bindings {
-                post: "/api/v1/namespaces/{namespace}/worker-deployment-versions/{deployment_version.deployment_name}/{deployment_version.build_id}/update-compute"
+                post: "/api/v1/namespaces/{namespace}/worker-deployment-versions/{deployment_version.deployment_name}/{deployment_version.build_id}/update-compute-config"
+                body: "*"
+            }
+        };
+    }
+
+    // Validates the compute config without attaching it to a Worker Deployment Version.
+    // Experimental. This API might significantly change or be removed in a future release.
+    rpc ValidateWorkerDeploymentVersionComputeConfig (ValidateWorkerDeploymentVersionComputeConfigRequest) returns (ValidateWorkerDeploymentVersionComputeConfigResponse) {
+        option (google.api.http) = {
+            post: "/namespaces/{namespace}/worker-deployment-versions/{deployment_version.deployment_name}/{deployment_version.build_id}/validate-compute-config"
+            body: "*"
+            additional_bindings {
+                post: "/api/v1/namespaces/{namespace}/worker-deployment-versions/{deployment_version.deployment_name}/{deployment_version.build_id}/validate-compute-config"
                 body: "*"
             }
         };