APPS-12973: Update spec to include request metrics for autoscaling and alerts

specification/resources/apps/models/app_alert_spec_rule.yml

@@ -16,5 +16,7 @@ enum:
   - FUNCTIONS_AVERAGE_WAIT_TIME_MS
   - FUNCTIONS_ERROR_COUNT
   - FUNCTIONS_GB_RATE_PER_SECOND
+  - REQUESTS_PER_SECOND
+  - REQUEST_DURATION_P95_MS
 type: string
 example: CPU_UTILIZATION

specification/resources/apps/models/app_autoscaling_spec.yml

@@ -0,0 +1,35 @@
+# CPU-only autoscaling for worker components. Services use app_autoscaling_spec_service.yml.
+description: Configuration for automatically scaling this component based on metrics.
+type: object
+properties:
+  min_instance_count:
+    description: The minimum amount of instances for this component.
+    type: integer
+    format: uint32
+    minimum: 1
+    example: 2
+  max_instance_count:
+    description: >-
+      The maximum amount of instances for this component. Maximum 250.
+      Consider using a larger instance size if your application requires more than 250 instances.
+    type: integer
+    format: uint32
+    minimum: 1
+    maximum: 250
+    example: 3
+  metrics:
+    description: The metrics that the component is scaled on.
+    type: object
+    properties:
+      cpu:
+        description: Settings for scaling the component based on CPU utilization.
+        type: object
+        properties:
+          percent:
+            description: The average target CPU utilization for the component.
+            type: integer
+            format: uint32
+            minimum: 1
+            maximum: 100
+            default: 80
+            example: 75

specification/resources/apps/models/app_autoscaling_spec_service.yml

@@ -0,0 +1,28 @@
+# Extends CPU autoscaling with request-based metrics. Only applicable to service components.
+allOf:
+  - $ref: app_autoscaling_spec.yml
+  - type: object
+    properties:
+      metrics:
+        type: object
+        properties:
+          requests_per_second:
+            description: Settings for scaling the component based on requests per second.
+            type: object
+            properties:
+              per_instance:
+                description: The target number of requests per second per instance for the component.
+                type: integer
+                format: uint32
+                minimum: 1
+                example: 100
+          request_duration:
+            description: Settings for scaling the component based on request duration.
+            type: object
+            properties:
+              p95_milliseconds:
+                description: The p95 target request duration in milliseconds for the component.
+                type: integer
+                format: uint32
+                minimum: 1
+                example: 500

specification/resources/apps/models/app_component_instance_base.yml

@@ -50,36 +50,3 @@ properties:
         example: basic-xxs
         deprecated: true
     example: apps-s-1vcpu-0.5gb
-
-  autoscaling:
-    description: Configuration for automatically scaling this component based on metrics.
-    type: object
-    properties:
-      min_instance_count:
-        description: The minimum amount of instances for this component. Must be less than max_instance_count.
-        type: integer
-        format: uint32
-        minimum: 1
-        example: 2
-      max_instance_count:
-        description: The maximum amount of instances for this component. Must be more than min_instance_count.
-        type: integer
-        format: uint32
-        minimum: 1
-        example: 3
-      metrics:
-        description: The metrics that the component is scaled on.
-        type: object
-        properties:
-          cpu:
-            description: Settings for scaling the component based on CPU utilization.
-            type: object
-            properties:
-              percent:
-                description: The average target CPU utilization for the component.
-                type: integer
-                format: uint32
-                minimum: 1
-                maximum: 100
-                default: 80
-                example: 75

specification/resources/apps/models/app_service_spec.yml

@@ -4,6 +4,9 @@ allOf:
 
 - type: object
   properties:
+    autoscaling:
+      $ref: app_autoscaling_spec_service.yml
+
     cors:
       allOf:
         - $ref: apps_cors_policy.yml

specification/resources/apps/models/app_worker_spec.yml

@@ -3,6 +3,9 @@ allOf:
 - $ref: app_component_instance_base.yml
 - type: object
   properties:
+    autoscaling:
+      $ref: app_autoscaling_spec.yml
+
     termination:
       $ref: app_worker_spec_termination.yml
     liveness_health_check: