fix(ecs): Improve health check and deployment configuration

- Increase HealthCheckGracePeriodSeconds from 0 to 300
- Add explicit DeploymentController type (ECS)
- Fix DeploymentConfiguration structure and add Alarms
- Update DeploymentCircuitBreaker configuration with proper Rollback handling

infra/cfn/forgejo-ecs-service.yaml

@@ -35,6 +35,7 @@ Parameters:
 
 Resources:
 
+
   ECSService:
     Type: AWS::ECS::Service
     Properties:
@@ -56,13 +57,18 @@ Resources:
         - TargetGroupArn: !ImportValue forgejo-network-TargetGroupArn
           ContainerName: forgejo
           ContainerPort: 3000
-      HealthCheckGracePeriodSeconds: 0
+      HealthCheckGracePeriodSeconds: 300
+      DeploymentController:
+        Type: ECS
       DeploymentConfiguration:
+        Strategy: ROLLING
         MaximumPercent: 100
         MinimumHealthyPercent: 0
-        DeploymentCircuitBreaker:
+        Alarms:
-          Enable: true
+          AlarmNames:
+            - myAlarm
           Rollback: true
+          Enable: true
       SchedulingStrategy: REPLICA
       EnableECSManagedTags: true
       PropagateTags: TASK_DEFINITION