fix: add task-level CPU and memory limits

terraform/modules/ecs-service/ecs_task_definition.tf

@@ -7,7 +7,10 @@ resource "aws_ecs_task_definition" "this" {
   network_mode          = var.network_mode
   pid_mode              = var.pid_mode
 
-  memory = var.container_memory_hard_limit
+  # TODO: Task-level cpu and memory values have been set using default/conservative values.
+  # They should be verified based on CloudWatch Container Insights or actual production metrics.
+  cpu    = var.task_cpu != null ? var.task_cpu : 1024
+  memory = var.task_memory != null ? var.task_memory : var.container_memory_hard_limit
 
   dynamic "placement_constraints" {
     for_each = var.placement_constraints

terraform/modules/ecs-service/variables.tf

@@ -452,3 +452,15 @@ variable "use_load_balancer" {
   type    = bool
   default = false
 }
+
+variable "task_cpu" {
+  description = "The number of cpu units used by the task. If not specified, a conservative default of 1024 is used."
+  type        = number
+  default     = null
+}
+
+variable "task_memory" {
+  description = "The amount (in MiB) of memory used by the task. If not specified, defaults to container_memory_hard_limit for backward compatibility."
+  type        = number
+  default     = null
+}