feat: Increase ECS autoscaling capacity (#907)

* feat: Increase ECS autoscaling capacity

* chore: Add changeset

* fix: Cpu low alarm and scaling adjustment as var

* feat: Set ecs min and max capacity switch region workflow

* fix: Set correct ecs scaling values

---------

Co-authored-by: GNuccio96 <giuseppenuccio.crea@pagopa.it>

Author: giuseppe-gangemi

.changeset/quiet-birds-wait.md

@@ -0,0 +1,6 @@
+---
+"infra": minor
+"oneidentity": minor
+---
+
+Increased autoscaling capacity of ecs-core tasks for bonus elettrodomestici

.github/workflows/switch-region.yml

@@ -28,7 +28,7 @@ jobs:
          --service-namespace ecs \
          --scalable-dimension ecs:service:DesiredCount \
          --resource-id service/$ECS_CLUSTER_NAME/$ECS_SERVICE_NAME \
-         --min-capacity 3 \
+         --min-capacity 60 --max-capacity 1000 \
          --region eu-central-1
 
 
@@ -38,7 +38,7 @@ jobs:
          aws ecs update-service \
          --cluster $ECS_CLUSTER_NAME \
          --service $ECS_SERVICE_NAME \
-         --desired-count 3 \
+         --desired-count 60 \
          --region eu-central-1
 
       - name: Wait for ECS Service to stabilize

src/infra/dev/eu-south-1/main.tf

@@ -195,10 +195,10 @@ module "backend" {
   vpc_cidr_block  = module.network.vpc_cidr_block
 
   service_core = {
-    service_name = format("%s-core", local.project)
-
-    cpu    = var.ecs_oneid_core.cpu
-    memory = var.ecs_oneid_core.memory
+    service_name                = format("%s-core", local.project)
+    cpu                         = var.ecs_oneid_core.cpu
+    memory                      = var.ecs_oneid_core.memory
+    cpu_high_scaling_adjustment = 2
 
     container = {
       name                = "oneid-core"

src/infra/dev/eu-south-1/variables.tf

@@ -475,8 +475,8 @@ variable "ecs_alarms" {
       namespace           = "AWS/ECS"
       evaluation_periods  = 3
       comparison_operator = "LessThanOrEqualToThreshold"
-      threshold           = 20
-      period              = 900
+      threshold           = 5
+      period              = 300
       statistic           = "Average"
       scaling_policy      = "cpu_low"
     },

src/infra/modules/backend/ecs.tf

@@ -426,7 +426,7 @@ module "ecs_core_service" {
         "adjustment_type" : "ChangeInCapacity"
         "step_adjustment" : [
           {
-            "scaling_adjustment" : 2 # Add 2 tasks
+            "scaling_adjustment" : var.service_core.cpu_high_scaling_adjustment # Add 10 tasks
             metric_interval_lower_bound = 0
           }
         ]
@@ -439,11 +439,11 @@ module "ecs_core_service" {
         "adjustment_type" : "ChangeInCapacity"
         "step_adjustment" : [
           {
-            "scaling_adjustment" : -1 # Add 2 tasks
+            "scaling_adjustment" : -1 # Remove 1 task
             metric_interval_lower_bound = 0
           }
         ]
-        cooldown = 900
+        cooldown = 300
       }
     }
   }

src/infra/modules/backend/variables.tf

@@ -66,10 +66,11 @@ variable "fargate_capacity_providers" {
 
 variable "service_core" {
   type = object({
-    service_name           = string
-    cpu                    = number
-    memory                 = number
-    enable_execute_command = optional(bool, true)
+    service_name                = string
+    cpu                         = number
+    memory                      = number
+    enable_execute_command      = optional(bool, true)
+    cpu_high_scaling_adjustment = number
     container = object({
       name                = string
       cpu                 = number

src/infra/prod/eu-central-1/README.md

@@ -131,7 +131,7 @@ No outputs.
 | <a name="input_dns_record_ttl"></a> [dns\_record\_ttl](#input\_dns\_record\_ttl) | Dns record ttl (in sec) | `number` | `86400` | no |
 | <a name="input_ecs_alarms"></a> [ecs\_alarms](#input\_ecs\_alarms) | n/a | <pre>map(object({<br/>    metric_name         = string<br/>    namespace           = string<br/>    threshold           = optional(number)<br/>    evaluation_periods  = optional(number)<br/>    period              = optional(number)<br/>    statistic           = optional(string)<br/>    comparison_operator = optional(string)<br/>    scaling_policy      = optional(string, null)<br/>  }))</pre> | <pre>{<br/>  "cpu_high": {<br/>    "comparison_operator": "GreaterThanOrEqualToThreshold",<br/>    "evaluation_periods": 1,<br/>    "metric_name": "CPUUtilization",<br/>    "namespace": "AWS/ECS",<br/>    "period": 60,<br/>    "scaling_policy": "cpu_high",<br/>    "statistic": "Average",<br/>    "threshold": 50<br/>  },<br/>  "cpu_low": {<br/>    "comparison_operator": "LessThanOrEqualToThreshold",<br/>    "evaluation_periods": 3,<br/>    "metric_name": "CPUUtilization",<br/>    "namespace": "AWS/ECS",<br/>    "period": 900,<br/>    "scaling_policy": "cpu_low",<br/>    "statistic": "Average",<br/>    "threshold": 20<br/>  },<br/>  "mem_high": {<br/>    "comparison_operator": "GreaterThanOrEqualToThreshold",<br/>    "evaluation_periods": 1,<br/>    "metric_name": "MemoryUtilization",<br/>    "namespace": "AWS/ECS",<br/>    "period": 60,<br/>    "statistic": "Average",<br/>    "threshold": 70<br/>  }<br/>}</pre> | no |
 | <a name="input_ecs_enable_container_insights"></a> [ecs\_enable\_container\_insights](#input\_ecs\_enable\_container\_insights) | Enable ecs cluster container inight. | `bool` | `true` | no |
-| <a name="input_ecs_oneid_core"></a> [ecs\_oneid\_core](#input\_ecs\_oneid\_core) | Oneidentity core backend configurations. | <pre>object({<br/>    image_version    = string<br/>    cpu              = number<br/>    memory           = number<br/>    container_cpu    = number<br/>    container_memory = number<br/>    autoscaling = object({<br/>      enable        = bool<br/>      desired_count = number<br/>      min_capacity  = number<br/>      max_capacity  = number<br/>    })<br/>    logs_retention_days   = number<br/>    app_spid_test_enabled = optional(bool, false)<br/>  })</pre> | <pre>{<br/>  "autoscaling": {<br/>    "desired_count": 1,<br/>    "enable": true,<br/>    "max_capacity": 12,<br/>    "min_capacity": 1<br/>  },<br/>  "container_cpu": 2048,<br/>  "container_memory": 4096,<br/>  "cpu": 2048,<br/>  "image_version": "05d9825a488fc80c2b8cd08d69e55732d4762977",<br/>  "logs_retention_days": 30,<br/>  "memory": 4096<br/>}</pre> | no |
+| <a name="input_ecs_oneid_core"></a> [ecs\_oneid\_core](#input\_ecs\_oneid\_core) | Oneidentity core backend configurations. | <pre>object({<br/>    image_version    = string<br/>    cpu              = number<br/>    memory           = number<br/>    container_cpu    = number<br/>    container_memory = number<br/>    autoscaling = object({<br/>      enable        = bool<br/>      desired_count = number<br/>      min_capacity  = number<br/>      max_capacity  = number<br/>    })<br/>    logs_retention_days   = number<br/>    app_spid_test_enabled = optional(bool, false)<br/>  })</pre> | <pre>{<br/>  "autoscaling": {<br/>    "desired_count": 40,<br/>    "enable": true,<br/>    "max_capacity": 80,<br/>    "min_capacity": 40<br/>  },<br/>  "container_cpu": 2048,<br/>  "container_memory": 4096,<br/>  "cpu": 2048,<br/>  "image_version": "05d9825a488fc80c2b8cd08d69e55732d4762977",<br/>  "logs_retention_days": 30,<br/>  "memory": 4096<br/>}</pre> | no |
 | <a name="input_enable_nat_gateway"></a> [enable\_nat\_gateway](#input\_enable\_nat\_gateway) | Create nat gateway(s) | `bool` | `false` | no |
 | <a name="input_env_short"></a> [env\_short](#input\_env\_short) | Environment short. | `string` | `"p"` | no |
 | <a name="input_idp_status_history_table"></a> [idp\_status\_history\_table](#input\_idp\_status\_history\_table) | IDP Status History configurations table. | <pre>object({<br/>    point_in_time_recovery_enabled = optional(bool, false)<br/>  })</pre> | <pre>{<br/>  "point_in_time_recovery_enabled": false<br/>}</pre> | no |

src/infra/prod/eu-central-1/main.tf

@@ -111,10 +111,10 @@ module "backend" {
   vpc_cidr_block  = module.network.vpc_cidr_block
 
   service_core = {
-    service_name = format("%s-core", local.project)
-
-    cpu    = var.ecs_oneid_core.cpu
-    memory = var.ecs_oneid_core.memory
+    service_name                = format("%s-core", local.project)
+    cpu_high_scaling_adjustment = 10
+    cpu                         = var.ecs_oneid_core.cpu
+    memory                      = var.ecs_oneid_core.memory
 
     container = {
       name                = "oneid-core"

src/infra/prod/eu-central-1/variables.tf

@@ -115,8 +115,8 @@ variable "ecs_oneid_core" {
     container_memory = 4096
     autoscaling = {
       enable        = true
-      desired_count = 1
-      min_capacity  = 1
+      desired_count = 3
+      min_capacity  = 3
       max_capacity  = 12
     }
     logs_retention_days = 30
@@ -423,8 +423,8 @@ variable "ecs_alarms" {
       namespace           = "AWS/ECS"
       evaluation_periods  = 3
       comparison_operator = "LessThanOrEqualToThreshold"
-      threshold           = 20
-      period              = 900
+      threshold           = 5
+      period              = 300
       statistic           = "Average"
       scaling_policy      = "cpu_low"
     },

src/infra/prod/eu-south-1/README.md

@@ -154,7 +154,7 @@
 | <a name="input_dns_record_ttl"></a> [dns\_record\_ttl](#input\_dns\_record\_ttl) | Dns record ttl (in sec) | `number` | `86400` | no |
 | <a name="input_ecs_alarms"></a> [ecs\_alarms](#input\_ecs\_alarms) | n/a | <pre>map(object({<br/>    metric_name         = string<br/>    namespace           = string<br/>    threshold           = optional(number)<br/>    evaluation_periods  = optional(number)<br/>    period              = optional(number)<br/>    statistic           = optional(string)<br/>    comparison_operator = optional(string)<br/>    scaling_policy      = optional(string)<br/>  }))</pre> | <pre>{<br/>  "cpu_high": {<br/>    "comparison_operator": "GreaterThanOrEqualToThreshold",<br/>    "evaluation_periods": 1,<br/>    "metric_name": "CPUUtilization",<br/>    "namespace": "AWS/ECS",<br/>    "period": 60,<br/>    "scaling_policy": "cpu_high",<br/>    "statistic": "Average",<br/>    "threshold": 50<br/>  },<br/>  "cpu_low": {<br/>    "comparison_operator": "LessThanOrEqualToThreshold",<br/>    "evaluation_periods": 3,<br/>    "metric_name": "CPUUtilization",<br/>    "namespace": "AWS/ECS",<br/>    "period": 900,<br/>    "scaling_policy": "cpu_low",<br/>    "statistic": "Average",<br/>    "threshold": 20<br/>  },<br/>  "mem_high": {<br/>    "comparison_operator": "GreaterThanOrEqualToThreshold",<br/>    "evaluation_periods": 1,<br/>    "metric_name": "MemoryUtilization",<br/>    "namespace": "AWS/ECS",<br/>    "period": 60,<br/>    "statistic": "Average",<br/>    "threshold": 70<br/>  }<br/>}</pre> | no |
 | <a name="input_ecs_enable_container_insights"></a> [ecs\_enable\_container\_insights](#input\_ecs\_enable\_container\_insights) | Enable ecs cluster container inight. | `bool` | `true` | no |
-| <a name="input_ecs_oneid_core"></a> [ecs\_oneid\_core](#input\_ecs\_oneid\_core) | Oneidentity core backend configurations. | <pre>object({<br/>    image_version    = string<br/>    cpu              = number<br/>    memory           = number<br/>    container_cpu    = number<br/>    container_memory = number<br/>    autoscaling = object({<br/>      enable        = bool<br/>      desired_count = number<br/>      min_capacity  = number<br/>      max_capacity  = number<br/>    })<br/>    logs_retention_days   = number<br/>    app_spid_test_enabled = optional(bool, false)<br/>  })</pre> | <pre>{<br/>  "autoscaling": {<br/>    "desired_count": 3,<br/>    "enable": true,<br/>    "max_capacity": 12,<br/>    "min_capacity": 3<br/>  },<br/>  "container_cpu": 2048,<br/>  "container_memory": 4096,<br/>  "cpu": 2048,<br/>  "image_version": "05d9825a488fc80c2b8cd08d69e55732d4762977",<br/>  "logs_retention_days": 30,<br/>  "memory": 4096<br/>}</pre> | no |
+| <a name="input_ecs_oneid_core"></a> [ecs\_oneid\_core](#input\_ecs\_oneid\_core) | Oneidentity core backend configurations. | <pre>object({<br/>    image_version    = string<br/>    cpu              = number<br/>    memory           = number<br/>    container_cpu    = number<br/>    container_memory = number<br/>    autoscaling = object({<br/>      enable        = bool<br/>      desired_count = number<br/>      min_capacity  = number<br/>      max_capacity  = number<br/>    })<br/>    logs_retention_days   = number<br/>    app_spid_test_enabled = optional(bool, false)<br/>  })</pre> | <pre>{<br/>  "autoscaling": {<br/>    "desired_count": 40,<br/>    "enable": true,<br/>    "max_capacity": 80,<br/>    "min_capacity": 40<br/>  },<br/>  "container_cpu": 2048,<br/>  "container_memory": 4096,<br/>  "cpu": 2048,<br/>  "image_version": "05d9825a488fc80c2b8cd08d69e55732d4762977",<br/>  "logs_retention_days": 30,<br/>  "memory": 4096<br/>}</pre> | no |
 | <a name="input_enable_nat_gateway"></a> [enable\_nat\_gateway](#input\_enable\_nat\_gateway) | Create nat gateway(s) | `bool` | `false` | no |
 | <a name="input_env_short"></a> [env\_short](#input\_env\_short) | Environment short. | `string` | `"p"` | no |
 | <a name="input_idp_metadata_table"></a> [idp\_metadata\_table](#input\_idp\_metadata\_table) | IDP Metadata configurations table. | <pre>object({<br/>    point_in_time_recovery_enabled = optional(bool, false)<br/>    stream_enabled                 = optional(bool, false)<br/>    stream_view_type               = optional(string, null)<br/>    deletion_protection_enabled    = optional(bool, false)<br/>    replication_regions = optional(list(object({<br/>      region_name            = string<br/>      propagate_tags         = optional(bool, true)<br/>      point_in_time_recovery = optional(bool, true)<br/>    })), [])<br/>  })</pre> | <pre>{<br/>  "deletion_protection_enabled": true,<br/>  "point_in_time_recovery_enabled": true,<br/>  "replication_regions": [<br/>    {<br/>      "propagate_tags": true,<br/>      "region_name": "eu-central-1"<br/>    }<br/>  ],<br/>  "stream_enabled": true,<br/>  "stream_view_type": "NEW_AND_OLD_IMAGES"<br/>}</pre> | no |