VED-1170: CloudWatch Alarm without actions

infrastructure/account/fhir_api_perf_alerts_slack_chatbot.tf

@@ -0,0 +1,24 @@
+resource "aws_chatbot_slack_channel_configuration" "fhir_api_perf_alerts" {
+  configuration_name = "${var.environment}-fhir-api-perf-alerts-slack-channel-config"
+  iam_role_arn       = aws_iam_role.fhir_api_perf_alerts_chatbot.arn
+  slack_channel_id   = var.environment == "prod" ? "C0B11MJPQ6A" : "C0B1GKZ5S4R"
+  slack_team_id      = "TJ00QR03U"
+  sns_topic_arns     = [aws_sns_topic.fhir_api_perf_alerts.arn]
+}
+
+resource "aws_iam_role" "fhir_api_perf_alerts_chatbot" {
+  name = "${var.environment}-fhir-api-perf-alerts-chatbot-channel-role"
+  assume_role_policy = jsonencode({
+    Version = "2012-10-17"
+    Statement = [
+      {
+        Action = "sts:AssumeRole"
+        Effect = "Allow"
+        Sid    = "AssumeChatbotRole"
+        Principal = {
+          Service = "chatbot.amazonaws.com"
+        }
+      },
+    ]
+  })
+}

infrastructure/account/fhir_api_perf_alerts_sns_topic.tf

@@ -0,0 +1,22 @@
+resource "aws_sns_topic" "fhir_api_perf_alerts" {
+  name              = "${var.environment}-fhir-api-perf-alerts"
+  kms_master_key_id = aws_kms_key.error_alerts_sns_encryption_key.arn
+}
+
+resource "aws_sns_topic_policy" "fhir_api_perf_alerts_topic_policy" {
+  arn = aws_sns_topic.fhir_api_perf_alerts.arn
+  policy = jsonencode({
+    Version = "2012-10-17",
+    Statement = [
+      {
+        Sid    = "AllowCloudWatchToPublish",
+        Effect = "Allow",
+        Principal = {
+          Service = "cloudwatch.amazonaws.com"
+        },
+        Action   = "SNS:Publish",
+        Resource = aws_sns_topic.fhir_api_perf_alerts.arn
+      }
+    ]
+  })
+}

infrastructure/account/kms.tf

@@ -225,3 +225,9 @@ resource "aws_kms_alias" "fhir_api_errors_sns_encryption_key" {
   name          = "alias/${var.environment}-fhir-api-errors-imms-sns-encryption"
   target_key_id = aws_kms_key.error_alerts_sns_encryption_key.key_id
 }
+
+resource "aws_kms_alias" "fhir_api_perf_alerts_sns_encryption_key" {
+  name          = "alias/${var.environment}-fhir-api-perf-alerts-imms-sns-encryption"
+  target_key_id = aws_kms_key.error_alerts_sns_encryption_key.key_id
+}
+

infrastructure/account/shield_protection.tf

@@ -34,13 +34,18 @@ locals {
   }
 }
 
+# Topic to publish alerts to when alarm is triggered
+data "aws_sns_topic" "fhir_api_perf_alerts" {
+  name = "${var.environment}-fhir-api-perf-alerts"
+}
 
 # Create Metric Alarms for each of those resources
 resource "aws_cloudwatch_metric_alarm" "ddos_protection_regional" {
   for_each = local.regional_shield_arn
 
   alarm_name        = "imms-${var.environment}-shield_ddos_${each.key}"
   alarm_description = "Alarm when Shield detects DDoS on ${each.key}"
+  alarm_actions     = [data.aws_sns_topic.fhir_api_perf_alerts.arn]
 
   namespace           = "AWS/DDoSProtection"
   metric_name         = "DDoSDetected"

infrastructure/instance/modules/lambda/lambda.tf

@@ -18,24 +18,28 @@ module "lambda_function_container_image" {
 
   # A JWT encode took 7 seconds at default memory size of 128 and 0.8 seconds at 1024.
   # 2048 gets it down to around 0.5 but since Lambda is charged at GB * ms then it costs more for minimal benefit.
-  memory_size = 1024
+  memory_size = var.memory_size
 
   environment_variables = var.environment_variables
   image_config_command  = ["${var.function_name}_handler.${var.function_name}_handler"]
 }
 
-resource "aws_cloudwatch_metric_alarm" "memory_alarm" {
-  alarm_name                = "${var.short_prefix}_${var.function_name} memory alarm"
-  comparison_operator       = "GreaterThanOrEqualToThreshold"
-  evaluation_periods        = 1
-  metric_name               = aws_cloudwatch_log_metric_filter.max_memory_used_metric.metric_transformation[0].name
-  namespace                 = aws_cloudwatch_log_metric_filter.max_memory_used_metric.metric_transformation[0].namespace
-  period                    = 600
-  statistic                 = "Maximum"
-  threshold                 = 256
-  alarm_description         = "This metric monitors Lambda memory usage"
-  insufficient_data_actions = []
+data "aws_sns_topic" "fhir_api_perf_alerts" {
+  name = "${var.environment}-fhir-api-perf-alerts"
+}
 
+resource "aws_cloudwatch_metric_alarm" "memory_alarm" {
+  alarm_name          = "${var.short_prefix}_${var.function_name} memory alarm"
+  comparison_operator = "GreaterThanOrEqualToThreshold"
+  evaluation_periods  = 1
+  metric_name         = aws_cloudwatch_log_metric_filter.max_memory_used_metric.metric_transformation[0].name
+  namespace           = aws_cloudwatch_log_metric_filter.max_memory_used_metric.metric_transformation[0].namespace
+  period              = 600
+  statistic           = "Maximum"
+  threshold           = var.memory_size * 0.85 # Alarm threshold set at 85% of memory size
+  alarm_description   = "This metric monitors Lambda memory usage"
+  alarm_actions       = [data.aws_sns_topic.fhir_api_perf_alerts.arn]
+  treat_missing_data  = "notBreaching"
 }
 
 resource "aws_cloudwatch_log_metric_filter" "max_memory_used_metric" {

infrastructure/instance/modules/lambda/variables.tf

@@ -41,4 +41,10 @@ variable "vpc_subnet_ids" {
 variable "environment" {
   description = "The deployment environment (e.g., dev, int, internal-qa, prod)"
   type        = string
+}
+
+variable "memory_size" {
+  description = "The memory size allocated to the lambda"
+  type        = number
+  default     = 1024
 }