Merge branch 'MON-521-monitors-for-aws-ecs-fargate' into 'master'

[MON-521] Monitors for AWS Fargate

Closes MON-521

See merge request claranet/pt-monitoring/projects/datadog/terraform/monitors!154

README.md

@@ -141,6 +141,7 @@ module "datadog-monitors-system-generic" {
 		- [ecs](https://github.com/claranet/terraform-datadog-monitors/tree/master/cloud/aws/ecs/)
 			- [common](https://github.com/claranet/terraform-datadog-monitors/tree/master/cloud/aws/ecs/common/)
 			- [ec2-cluster](https://github.com/claranet/terraform-datadog-monitors/tree/master/cloud/aws/ecs/ec2-cluster/)
+			- [fargate](https://github.com/claranet/terraform-datadog-monitors/tree/master/cloud/aws/ecs/fargate/)
 		- [elasticache](https://github.com/claranet/terraform-datadog-monitors/tree/master/cloud/aws/elasticache/)
 			- [common](https://github.com/claranet/terraform-datadog-monitors/tree/master/cloud/aws/elasticache/common/)
 			- [memcached](https://github.com/claranet/terraform-datadog-monitors/tree/master/cloud/aws/elasticache/memcached/)

cloud/aws/ecs/fargate/MANIFEST.txt

@@ -0,0 +1 @@
+cloud-aws

cloud/aws/ecs/fargate/README.md

@@ -0,0 +1,73 @@
+# CLOUD AWS ECS FARGATE DataDog monitors
+
+## How to use this module
+
+```hcl
+module "datadog-monitors-cloud-aws-ecs-fargate" {
+  source      = "claranet/monitors/datadog//cloud/aws/ecs/fargate"
+  version     = "{revision}"
+
+  environment = var.environment
+  message     = module.datadog-message-alerting.alerting-message
+}
+
+```
+
+## Purpose
+
+Creates DataDog monitors with the following checks:
+
+- Fargate CPU Utilization High (disabled by default)
+- Fargate memory Utilization High (disabled by default)
+- Fargate service does not respond.
+
+## Inputs
+
+| Name | Description | Type | Default | Required |
+|------|-------------|------|---------|:-----:|
+| cpu\_utilization\_enabled | Flag to enable monitor | `string` | `"false"` | no |
+| cpu\_utilization\_extra\_tags | Extra tags for the monitor | `list(string)` | `[]` | no |
+| cpu\_utilization\_message | Custom message for the monitor | `string` | `""` | no |
+| cpu\_utilization\_threshold\_critical | Critical threshold for the monitor | `string` | `90` | no |
+| cpu\_utilization\_threshold\_warning | Warning threshold for the monitor | `string` | `85` | no |
+| cpu\_utilization\_time\_aggregator | Monitor aggregator (min, max or avg) | `string` | `"min"` | no |
+| cpu\_utilization\_timeframe | Timeframe for the monitor | `string` | `"last_5m"` | no |
+| environment | Architecture environment | `string` | n/a | yes |
+| evaluation\_delay | Delay in seconds for the metric evaluation | `number` | `15` | no |
+| filter\_tags\_custom | Tags used for custom filtering when filter\_tags\_use\_defaults is false | `string` | `"*"` | no |
+| filter\_tags\_custom\_excluded | Tags excluded for custom filtering when filter\_tags\_use\_defaults is false | `string` | `""` | no |
+| filter\_tags\_use\_defaults | Use default filter tags convention | `bool` | `true` | no |
+| memory\_utilization\_enabled | Flag to enable Fargate Memory utilization monitor | `string` | `"false"` | no |
+| memory\_utilization\_extra\_tags | Extra tags for Fargate Memory utilization monitor | `list(string)` | `[]` | no |
+| memory\_utilization\_message | Custom message for the Fargate Memory Utilization monitor | `string` | `""` | no |
+| memory\_utilization\_threshold\_critical | Critical threshold for the Fargate Memory Utilization monitor | `string` | `90` | no |
+| memory\_utilization\_threshold\_warning | Warning threshold for the Fargate Memory Utilization monitor | `string` | `85` | no |
+| memory\_utilization\_time\_aggregator | Monitor aggregator for Fargate Memory Utilization [available values: min, max or avg] | `string` | `"min"` | no |
+| memory\_utilization\_timeframe | Timeframe for the Fargate Memory Utilization monitor | `string` | `"last_5m"` | no |
+| message | Message sent when a monitor is triggered | `string` | n/a | yes |
+| new\_host\_delay | Delay in seconds before monitor new resource | `number` | `300` | no |
+| notify\_no\_data | Will raise no data alert if set to true | `bool` | `true` | no |
+| prefix\_slug | Prefix string to prepend between brackets on every monitors names | `string` | `""` | no |
+| service\_check\_enabled | Flag to enable monitor | `bool` | `true` | no |
+| service\_check\_extra\_tags | Extra tags for the monitor | `list(string)` | `[]` | no |
+| service\_check\_message | Custom message for the monitor | `string` | `""` | no |
+| service\_check\_no\_data\_timeframe | No data timeframe in minutes | `number` | `10` | no |
+| service\_check\_threshold\_warning | Warning threshold | `number` | `3` | no |
+
+## Outputs
+
+| Name | Description |
+|------|-------------|
+| cpu\_utilization\_id | id for monitor cpu\_utilization |
+| memory\_utilization\_id | id for monitor memory\_utilization |
+| service\_check\_id | id for monitor service\_check |
+
+## Related documentation
+
+[Official DataDog documentation on ECS Fargate](https://docs.datadoghq.com/integrations/ecs_fargate/)
+
+### Specific configuration due to agent limitations
+
+CPU & memory monitors will be usable only when deploying datadog agent as a sidecar in task definitions.
+
+In order to avoid clutter on monitors, datadog agent & ECS internal containers are always excluded from filtering to be on par with Kubernetes way of work. A bug is [currently opened](https://github.com/DataDog/datadog-agent/issues/2722) on agent repository on this matter.

cloud/aws/ecs/fargate/inputs.tf

@@ -0,0 +1,169 @@
+# Generics
+
+variable "environment" {
+  description = "Architecture environment"
+  type        = string
+}
+
+variable "message" {
+  type        = string
+  description = "Message sent when a monitor is triggered"
+}
+
+variable "evaluation_delay" {
+  description = "Delay in seconds for the metric evaluation"
+  type        = number
+  default     = 15
+}
+
+variable "new_host_delay" {
+  description = "Delay in seconds before monitor new resource"
+  type        = number
+  default     = 300
+}
+
+variable "notify_no_data" {
+  description = "Will raise no data alert if set to true"
+  type        = bool
+  default     = true
+}
+
+variable "prefix_slug" {
+  description = "Prefix string to prepend between brackets on every monitors names"
+  default     = ""
+}
+
+variable "filter_tags_use_defaults" {
+  description = "Use default filter tags convention"
+  type        = bool
+  default     = true
+}
+
+variable "filter_tags_custom" {
+  description = "Tags used for custom filtering when filter_tags_use_defaults is false"
+  type        = string
+  default     = "*"
+}
+
+variable "filter_tags_custom_excluded" {
+  description = "Tags excluded for custom filtering when filter_tags_use_defaults is false"
+  type        = string
+  default     = ""
+}
+
+# Service checks
+variable "service_check_enabled" {
+  type        = bool
+  description = "Flag to enable monitor "
+  default     = true
+}
+
+variable "service_check_message" {
+  type        = string
+  description = "Custom message for the monitor"
+  default     = ""
+}
+
+variable "service_check_extra_tags" {
+  type        = list(string)
+  description = "Extra tags for the monitor"
+  default     = []
+}
+
+variable "service_check_threshold_warning" {
+  type        = number
+  description = "Warning threshold"
+  default     = 3
+}
+
+variable "service_check_no_data_timeframe" {
+  type        = number
+  description = "No data timeframe in minutes"
+  default     = 10
+}
+
+# CPU utilization
+variable "cpu_utilization_enabled" {
+  description = "Flag to enable monitor"
+  type        = string
+  default     = "false"
+}
+
+variable "cpu_utilization_message" {
+  description = "Custom message for the monitor"
+  type        = string
+  default     = ""
+}
+
+variable "cpu_utilization_time_aggregator" {
+  description = "Monitor aggregator (min, max or avg)"
+  type        = string
+  default     = "min"
+}
+
+variable "cpu_utilization_timeframe" {
+  description = "Timeframe for the monitor"
+  type        = string
+  default     = "last_5m"
+}
+
+variable "cpu_utilization_threshold_critical" {
+  description = "Critical threshold for the monitor"
+  type        = string
+  default     = 90
+}
+
+variable "cpu_utilization_threshold_warning" {
+  description = "Warning threshold for the monitor"
+  type        = string
+  default     = 85
+}
+
+variable "cpu_utilization_extra_tags" {
+  description = "Extra tags for the monitor"
+  type        = list(string)
+  default     = []
+}
+
+# Memory usage
+variable "memory_utilization_enabled" {
+  description = "Flag to enable Fargate Memory utilization monitor"
+  type        = string
+  default     = "false"
+}
+
+variable "memory_utilization_extra_tags" {
+  description = "Extra tags for Fargate Memory utilization monitor"
+  type        = list(string)
+  default     = []
+}
+
+variable "memory_utilization_message" {
+  description = "Custom message for the Fargate Memory Utilization monitor"
+  type        = string
+  default     = ""
+}
+
+variable "memory_utilization_time_aggregator" {
+  description = "Monitor aggregator for Fargate Memory Utilization [available values: min, max or avg]"
+  type        = string
+  default     = "min"
+}
+
+variable "memory_utilization_timeframe" {
+  description = "Timeframe for the Fargate Memory Utilization monitor"
+  type        = string
+  default     = "last_5m"
+}
+
+variable "memory_utilization_threshold_critical" {
+  description = "Critical threshold for the Fargate Memory Utilization monitor"
+  type        = string
+  default     = 90
+}
+
+variable "memory_utilization_threshold_warning" {
+  description = "Warning threshold for the Fargate Memory Utilization monitor"
+  type        = string
+  default     = 85
+}

cloud/aws/ecs/fargate/modules.tf

@@ -0,0 +1,14 @@
+module "filter-tags" {
+
+  source = "../../../../common/filter-tags"
+
+  environment = var.environment
+  resource    = "aws_ecs"
+  extra_tags_excluded = [
+    "ecs_container_name:datadog-agent",
+    "ecs_container_name:_internal_ecs_pause"
+  ]
+  filter_tags_use_defaults    = var.filter_tags_use_defaults
+  filter_tags_custom          = var.filter_tags_custom
+  filter_tags_custom_excluded = var.filter_tags_custom_excluded
+}

cloud/aws/ecs/fargate/monitors-ecs-fargate.tf

@@ -0,0 +1,110 @@
+# Service check
+
+resource "datadog_monitor" "service_check" {
+  count   = var.service_check_enabled ? 1 : 0
+  name    = "${var.prefix_slug == "" ? "" : "[${var.prefix_slug}]"}[${var.environment}] Fargate service does not respond."
+  message = coalesce(var.service_check_message, var.message)
+  type    = "service check"
+
+  query = <<EOQ
+"fargate_check"${module.filter-tags.service_check}.last(6).count_by_status()
+EOQ
+
+  thresholds = {
+    warning  = var.service_check_threshold_warning
+    critical = 5
+  }
+
+  notify_audit = false
+  locked       = false
+  timeout_h    = 0
+
+  evaluation_delay  = var.evaluation_delay
+  new_host_delay    = var.new_host_delay
+  notify_no_data    = var.notify_no_data
+  no_data_timeframe = var.service_check_no_data_timeframe
+  renotify_interval = 0
+
+
+  include_tags = true
+
+  tags = concat(["env:${var.environment}", "type:cloud", "provider:aws", "resource:ecs_fargate", "team:claranet", "created-by:terraform", "category:service"], var.service_check_extra_tags)
+
+  lifecycle {
+    ignore_changes = [silenced]
+  }
+}
+
+resource "datadog_monitor" "cpu_utilization" {
+  count   = var.cpu_utilization_enabled ? 1 : 0
+  name    = "${var.prefix_slug == "" ? "" : "[${var.prefix_slug}]"}[${var.environment}] Fargate CPU Utilization High {{#is_alert}}{{{comparator}}} {{threshold}}% ({{value}}%){{/is_alert}}{{#is_warning}}{{{comparator}}} {{warn_threshold}}% ({{value}}%){{/is_warning}}"
+  message = coalesce(var.cpu_utilization_message, var.message)
+  type    = "metric alert"
+
+  query = <<EOQ
+${var.cpu_utilization_time_aggregator}(${var.cpu_utilization_timeframe}):
+  avg:ecs.fargate.cpu.percent${module.filter-tags.query_alert} by {task_family, ecs_container_name}
+> ${var.cpu_utilization_threshold_critical}
+EOQ
+
+
+  thresholds = {
+    critical = var.cpu_utilization_threshold_critical
+    warning  = var.cpu_utilization_threshold_warning
+  }
+
+  evaluation_delay    = var.evaluation_delay
+  new_host_delay      = var.new_host_delay
+  notify_no_data      = false
+  require_full_window = true
+  renotify_interval   = 0
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+
+  tags = concat(["env:${var.environment}", "type:cloud", "provider:aws", "resource:ecs_fargate", "team:claranet", "created-by:terraform"], var.cpu_utilization_extra_tags)
+
+  lifecycle {
+    ignore_changes = [silenced]
+  }
+
+}
+
+resource "datadog_monitor" "memory_utilization" {
+  count   = var.memory_utilization_enabled == "true" ? 1 : 0
+  name    = "${var.prefix_slug == "" ? "" : "[${var.prefix_slug}]"}[${var.environment}] Fargate memory Utilization High {{#is_alert}}{{{comparator}}} {{threshold}}% ({{value}}%){{/is_alert}}{{#is_warning}}{{{comparator}}} {{warn_threshold}}% ({{value}}%){{/is_warning}}"
+  message = coalesce(var.memory_utilization_message, var.message)
+  type    = "metric alert"
+
+  query = <<EOQ
+${var.memory_utilization_time_aggregator}(${var.memory_utilization_timeframe}):
+  avg:ecs.fargate.mem.usage${module.filter-tags.query_alert} by {task_family, ecs_container_name} /
+  avg:ecs.fargate.mem.limit${module.filter-tags.query_alert} by {task_family, ecs_container_name}
+  * 100
+> ${var.memory_utilization_threshold_critical}
+EOQ
+
+
+  thresholds = {
+    critical = var.memory_utilization_threshold_critical
+    warning  = var.memory_utilization_threshold_warning
+  }
+
+  evaluation_delay    = var.evaluation_delay
+  new_host_delay      = var.new_host_delay
+  notify_no_data      = false
+  require_full_window = true
+  renotify_interval   = 0
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+
+  tags = concat(["env:${var.environment}", "type:cloud", "provider:aws", "resource:ecs_fargate", "team:claranet", "created-by:terraform"], var.memory_utilization_extra_tags)
+
+  lifecycle {
+    ignore_changes = [silenced]
+  }
+
+}

cloud/aws/ecs/fargate/outputs.tf

@@ -0,0 +1,15 @@
+output "cpu_utilization_id" {
+  description = "id for monitor cpu_utilization"
+  value       = datadog_monitor.cpu_utilization.*.id
+}
+
+output "memory_utilization_id" {
+  description = "id for monitor memory_utilization"
+  value       = datadog_monitor.memory_utilization.*.id
+}
+
+output "service_check_id" {
+  description = "id for monitor service_check"
+  value       = datadog_monitor.service_check.*.id
+}
+