CCM-21555: tweak alarm settings write throttling

infrastructure/terraform/components/api/module_ddb_alarms_letter_queue.tf

@@ -4,4 +4,7 @@ module "ddb_alarms_letter_queue" {
   alarm_prefix = local.csi
   table_name   = aws_dynamodb_table.letter_queue.name
   tags         = local.default_tags
+
+  write_throttle_evaluation_periods  = 5
+  write_throttle_datapoints_to_alarm = 3
 }

infrastructure/terraform/components/api/variables.tf

@@ -226,6 +226,7 @@ variable "enable_alarms" {
   description = "Enable CloudWatch alarms for this deployed environment"
   default     = true
 }
+
 variable "deploy_supplier_mock_scheduler" {
   type        = bool
   description = "Deploy EventBridge Scheduler trigger for supplier mock lambda"

infrastructure/terraform/modules/alarms-ddb/README.md

@@ -17,6 +17,9 @@
 | <a name="input_read_throttle_threshold"></a> [read\_throttle\_threshold](#input\_read\_throttle\_threshold) | n/a | `number` | `0` | no |
 | <a name="input_table_name"></a> [table\_name](#input\_table\_name) | n/a | `string` | n/a | yes |
 | <a name="input_tags"></a> [tags](#input\_tags) | n/a | `map(string)` | `{}` | no |
+| <a name="input_write_capacity_threshold"></a> [write\_capacity\_threshold](#input\_write\_capacity\_threshold) | n/a | `number` | `48000` | no |
+| <a name="input_write_throttle_datapoints_to_alarm"></a> [write\_throttle\_datapoints\_to\_alarm](#input\_write\_throttle\_datapoints\_to\_alarm) | n/a | `number` | `1` | no |
+| <a name="input_write_throttle_evaluation_periods"></a> [write\_throttle\_evaluation\_periods](#input\_write\_throttle\_evaluation\_periods) | n/a | `number` | `1` | no |
 | <a name="input_write_throttle_threshold"></a> [write\_throttle\_threshold](#input\_write\_throttle\_threshold) | n/a | `number` | `0` | no |
 ## Modules
 

infrastructure/terraform/modules/alarms-ddb/cloudwatch_metric_alarm_write_capacity.tf

@@ -0,0 +1,22 @@
+resource "aws_cloudwatch_metric_alarm" "write_capacity" {
+  alarm_name        = "${var.alarm_prefix}-ddb-${var.table_name}-write-capacity"
+  alarm_description = "RELIABILITY: DynamoDB consumed write capacity approaching limit"
+
+  namespace   = "AWS/DynamoDB"
+  metric_name = "ConsumedWriteCapacityUnits"
+  statistic   = "Sum"
+  period      = var.period_seconds
+
+  evaluation_periods  = var.evaluation_periods
+  threshold           = var.write_capacity_threshold
+  comparison_operator = "GreaterThanThreshold"
+  treat_missing_data  = "notBreaching"
+
+  dimensions = { TableName = var.table_name }
+
+  actions_enabled           = false
+  alarm_actions             = []
+  ok_actions                = []
+  insufficient_data_actions = []
+  tags                      = var.tags
+}

infrastructure/terraform/modules/alarms-ddb/cloudwatch_metric_alarm_write_throttle.tf

@@ -7,7 +7,8 @@ resource "aws_cloudwatch_metric_alarm" "write_throttle" {
   statistic   = "Sum"
   period      = var.period_seconds
 
-  evaluation_periods  = var.evaluation_periods
+  evaluation_periods  = var.write_throttle_evaluation_periods
+  datapoints_to_alarm = var.write_throttle_datapoints_to_alarm
   threshold           = var.write_throttle_threshold
   comparison_operator = "GreaterThanThreshold"
   treat_missing_data  = "notBreaching"

infrastructure/terraform/modules/alarms-ddb/variables.tf

@@ -21,6 +21,16 @@ variable "evaluation_periods" {
   default = 1
 }
 
+variable "write_throttle_evaluation_periods" {
+  type    = number
+  default = 1
+}
+
+variable "write_throttle_datapoints_to_alarm" {
+  type    = number
+  default = 1
+}
+
 variable "read_throttle_threshold" {
   type    = number
   default = 0
@@ -30,3 +40,8 @@ variable "write_throttle_threshold" {
   type    = number
   default = 0
 }
+
+variable "write_capacity_threshold" {
+  type    = number
+  default = 48000 # 80% of 60k/min DynamoDB limit
+}