Capacity provider feature (#22)

Signed-off-by: Arthur Diniz <[email protected]>

Author: arthurbdiniz

.github/workflows/lint.yml

@@ -51,7 +51,7 @@ jobs:
     name: Minimum version check
     runs-on: ubuntu-latest
     container:
-      image: hashicorp/terraform:0.12.0
+      image: hashicorp/terraform:0.13.0
     steps:
       - uses: actions/checkout@master
       - name: Validate Code

_variables.tf

@@ -1,19 +1,19 @@
 # == REQUIRED VARS
 
 variable "name" {
-  description = "Name of this ECS cluster"
+  description = "Name of this ECS cluster."
 }
 
 variable "instance_type_1" {
-  description = "Instance type for ECS workers (first priority)"
+  description = "Instance type for ECS workers (first priority)."
 }
 
 variable "instance_type_2" {
-  description = "Instance type for ECS workers (second priority)"
+  description = "Instance type for ECS workers (second priority)."
 }
 
 variable "instance_type_3" {
-  description = "Instance type for ECS workers (third priority)"
+  description = "Instance type for ECS workers (third priority)."
 }
 
 variable "architecture" {
@@ -22,7 +22,7 @@ variable "architecture" {
 }
 
 variable "on_demand_percentage" {
-  description = "Percentage of on-demand intances vs spot"
+  description = "Percentage of on-demand intances vs spot."
   default     = 100
 }
 
@@ -32,22 +32,22 @@ variable "on_demand_base_capacity" {
 }
 
 variable "vpc_id" {
-  description = "VPC ID to deploy the ECS cluster"
+  description = "VPC ID to deploy the ECS cluster."
 }
 
 variable "private_subnet_ids" {
   type        = list(string)
-  description = "List of private subnet IDs for ECS instances and Internal ALB when enabled"
+  description = "List of private subnet IDs for ECS instances and Internal ALB when enabled."
 }
 
 variable "public_subnet_ids" {
   type        = list(string)
-  description = "List of public subnet IDs for ECS ALB"
+  description = "List of public subnet IDs for ECS ALB."
 }
 
 variable "secure_subnet_ids" {
   type        = list(string)
-  description = "List of secure subnet IDs for EFS"
+  description = "List of secure subnet IDs for EFS."
 }
 
 variable "certificate_arn" {}
@@ -57,27 +57,27 @@ variable "certificate_arn" {}
 variable "security_group_ids" {
   type        = list(string)
   default     = []
-  description = "Extra security groups for instances"
+  description = "Extra security groups for instances."
 }
 
 variable "userdata" {
   default     = ""
-  description = "Extra commands to pass to userdata"
+  description = "Extra commands to pass to userdata."
 }
 
 variable "alb" {
   default     = true
-  description = "Whether to deploy an ALB or not with the cluster"
+  description = "Whether to deploy an ALB or not with the cluster."
 }
 
 variable "alb_only" {
   default     = false
-  description = "Whether to deploy only an alb and no cloudFront or not with the cluster"
+  description = "Whether to deploy only an alb and no cloudFront or not with the cluster."
 }
 
 variable "alb_internal" {
   default     = false
-  description = "Deploys a second internal ALB for private APIs"
+  description = "Deploys a second internal ALB for private APIs."
 }
 
 variable "certificate_internal_arn" {
@@ -93,129 +93,139 @@ variable "alb_ssl_policy" {
 
 variable "asg_min" {
   default     = 1
-  description = "Min number of instances for autoscaling group"
+  description = "Min number of instances for autoscaling group."
 }
 
 variable "asg_max" {
   default     = 4
-  description = "Max number of instances for autoscaling group"
+  description = "Max number of instances for autoscaling group."
 }
 
-variable "asg_memory_target" {
-  default     = 60
-  description = "Target average memory percentage to track for autoscaling"
+variable "asg_protect_from_scale_in" {
+  default     = false
+  description = "(Optional) Allows setting instance protection. The autoscaling group will not select instances with this setting for termination during scale in events."
+}
+
+variable "asg_target_capacity" {
+  default     = 70
+  description = "Target average capacity percentage for the ECS capacity provider to track for autoscaling."
 }
 
 variable "alarm_sns_topics" {
   default     = []
-  description = "Alarm topics to create and alert on ECS instance metrics"
+  description = "Alarm topics to create and alert on ECS instance metrics."
 }
 
 variable "alarm_asg_high_cpu_threshold" {
-  description = "Max threshold average CPU percentage allowed in a 2 minutes interval (use 0 to disable this alarm)"
+  description = "Max threshold average CPU percentage allowed in a 2 minutes interval (use 0 to disable this alarm)."
   default     = 80
 }
 
 variable "alarm_ecs_high_memory_threshold" {
-  description = "Max threshold average Memory percentage allowed in a 2 minutes interval (use 0 to disable this alarm)"
+  description = "Max threshold average Memory percentage allowed in a 2 minutes interval (use 0 to disable this alarm)."
   default     = 80
 }
 
 variable "alarm_ecs_high_cpu_threshold" {
-  description = "Max threshold average CPU percentage allowed in a 2 minutes interval (use 0 to disable this alarm)"
+  description = "Max threshold average CPU percentage allowed in a 2 minutes interval (use 0 to disable this alarm)."
   default     = 80
 }
 
 variable "alarm_alb_latency_anomaly_threshold" {
-  description = "ALB Latency anomaly detection width (use 0 to disable this alarm)"
+  description = "ALB Latency anomaly detection width (use 0 to disable this alarm)."
   default     = 2
 }
 
 variable "alarm_alb_500_errors_threshold" {
-  description = "Max threshold of HTTP 500 errors allowed in a 5 minutes interval (use 0 to disable this alarm)"
+  description = "Max threshold of HTTP 500 errors allowed in a 5 minutes interval (use 0 to disable this alarm)."
   default     = 10
 }
 
 variable "alarm_alb_400_errors_threshold" {
-  description = "Max threshold of HTTP 4000 errors allowed in a 5 minutes interval (use 0 to disable this alarm)"
+  description = "Max threshold of HTTP 4000 errors allowed in a 5 minutes interval (use 0 to disable this alarm)."
   default     = 10
 }
 
 variable "alarm_efs_credits_low_threshold" {
-  description = "Alerts when EFS credits fell below this number in bytes - default 1000000000000 is 1TB of a maximum of 2.31T of credits (use 0 to disable this alarm)"
+  description = "Alerts when EFS credits fell below this number in bytes - default 1000000000000 is 1TB of a maximum of 2.31T of credits (use 0 to disable this alarm)."
   default     = 1000000000000
 }
 
 variable "target_group_arns" {
   default     = []
   type        = list(string)
-  description = "List of target groups for ASG to register"
+  description = "List of target groups for ASG to register."
 }
 
 variable "autoscaling_health_check_grace_period" {
   default     = 300
-  description = "The length of time that Auto Scaling waits before checking an instance's health status. The grace period begins when an instance comes into service"
+  description = "The length of time that Auto Scaling waits before checking an instance's health status. The grace period begins when an instance comes into service."
 }
 
 variable "autoscaling_default_cooldown" {
   default     = 300
-  description = "The amount of time, in seconds, after a scaling activity completes before another scaling activity can start"
+  description = "The amount of time, in seconds, after a scaling activity completes before another scaling activity can start."
 }
 
 variable "instance_volume_size" {
-  description = "Volume size for docker volume (in GB)"
-  default     = 30
+  description = "Volume size for docker volume (in GB)."
+  default     = 22
+}
+
+variable "instance_volume_size_root" {
+  description = "Volume size for root volume (in GB)."
+  default     = 16
 }
 
 variable "lb_access_logs_bucket" {
   type        = string
   default     = ""
-  description = "Bucket to store logs from lb access"
+  description = "Bucket to store logs from lb access."
 }
 
 variable "lb_access_logs_prefix" {
   type        = string
   default     = ""
-  description = "Bucket prefix to store lb access logs"
+  description = "Bucket prefix to store lb access logs."
 }
 
 variable "enable_schedule" {
   default     = false
-  description = "Enables schedule to shut down and start up instances outside business hours"
+  description = "Enables schedule to shut down and start up instances outside business hours."
 }
 variable "schedule_cron_start" {
   type        = string
   default     = ""
-  description = "Cron expression to define when to trigger a start of the auto-scaling group. E.g. '0 20 * * *' to start at 8pm GMT time"
+  description = "Cron expression to define when to trigger a start of the auto-scaling group. E.g. '0 20 * * *' to start at 8pm GMT time."
 }
 
 variable "schedule_cron_stop" {
   type        = string
   default     = ""
-  description = "Cron expression to define when to trigger a stop of the auto-scaling group. E.g. '0 10 * * *' to stop at 10am GMT time"
+  description = "Cron expression to define when to trigger a stop of the auto-scaling group. E.g. '0 10 * * *' to stop at 10am GMT time."
 }
 
 variable "backup" {
   type        = string
   default     = "true"
-  description = "Assing a backup tag to efs resource - Backup will be performed by AWS Backup"
+  description = "Assing a backup tag to efs resource - Backup will be performed by AWS Backup."
 }
 
 variable "throughput_mode" {
   type        = string
   default     = "bursting"
-  description = "Throughput mode for the file system. Defaults to bursting. Valid values: bursting, provisioned"
+  description = "Throughput mode for the file system. Defaults to bursting. Valid values: bursting, provisioned."
 }
 
 variable "provisioned_throughput_in_mibps" {
   default     = 0
-  description = "The throughput, measured in MiB/s, that you want to provision for the file system"
+  description = "The throughput, measured in MiB/s, that you want to provision for the file system."
 }
 
 variable "alarm_prefix" {
   type        = string
   description = "String prefix for cloudwatch alarms. (Optional)"
-  default     = ""
+  default     = "alarm"
 }
 
 variable "kms_key_arn" {

asg.tf

@@ -33,6 +33,8 @@ resource "aws_autoscaling_group" "ecs" {
   min_size = var.asg_min
   max_size = var.asg_max
 
+  protect_from_scale_in = var.asg_protect_from_scale_in
+
   tags = [
     map("key", "Name", "value", "ecs-node-${var.name}", "propagate_at_launch", true)
   ]
@@ -45,25 +47,18 @@ resource "aws_autoscaling_group" "ecs" {
   }
 }
 
-resource "aws_autoscaling_policy" "ecs_memory_tracking" {
-  name                      = "ecs-${var.name}-memory"
-  policy_type               = "TargetTrackingScaling"
-  autoscaling_group_name    = aws_autoscaling_group.ecs.name
-  estimated_instance_warmup = "180"
+resource "aws_ecs_capacity_provider" "ecs_capacity_provider" {
+  name = "${var.name}-capacity-provider"
 
-  target_tracking_configuration {
-    customized_metric_specification {
-      metric_dimension {
-        name  = "ClusterName"
-        value = aws_ecs_cluster.ecs.name
-      }
+  auto_scaling_group_provider {
+    auto_scaling_group_arn         = aws_autoscaling_group.ecs.arn
+    managed_termination_protection = "DISABLED"
 
-      metric_name = "MemoryReservation"
-      namespace   = "AWS/ECS"
-      statistic   = "Average"
-      unit        = "Percent"
+    managed_scaling {
+      maximum_scaling_step_size = 10
+      minimum_scaling_step_size = 1
+      status                    = "ENABLED"
+      target_capacity           = var.asg_target_capacity
     }
-
-    target_value = var.asg_memory_target
   }
 }

ec2-launch-template.tf

@@ -2,7 +2,7 @@ data "template_file" "userdata" {
   template = file("${path.module}/userdata.tpl")
 
   vars = {
-    tf_cluster_name = aws_ecs_cluster.ecs.name
+    tf_cluster_name = var.name
     tf_efs_id       = aws_efs_file_system.ecs.id
     userdata_extra  = var.userdata
   }

ecs.tf

@@ -1,5 +1,9 @@
 resource "aws_ecs_cluster" "ecs" {
-  name = var.name
+
+  depends_on = [aws_autoscaling_group.ecs, aws_ecs_capacity_provider.ecs_capacity_provider]
+  name       = var.name
+
+  capacity_providers = ["${var.name}-capacity-provider"]
 
   lifecycle {
     ignore_changes = [

versions.tf

@@ -1,3 +1,3 @@
 terraform {
-  required_version = ">= 0.12.0"
+  required_version = ">= 0.13.0"
 }