MON-90 Azure API Management monitors

cloud/azure/apimanagement/README.md

@@ -0,0 +1,43 @@
+Azure API Management Datadog monitors
+=====================================
+
+How to use this module
+----------------------
+```
+module "datadog-monitors-azure-apimanagement" {
+  source = "git::ssh://git@bitbucket.org/morea/terraform.feature.datadog.git//cloud/azure/apimanagement?ref={revision}"
+
+  message     = "${module.datadog-message-alerting.alerting-message}"
+  environment = "${var.environment}"
+}
+```
+
+Purpose
+-------
+Creates Datadog monitors with the following checks :
+
+* Service status
+* Failed requests ratio
+* Other requests ratio
+* Unauthorized requests ratio
+* Successful requests ratio
+
+Inputs
+------
+
+| Name | Description | Type | Default | Required |
+|------|-------------|:----:|:-----:|:-----:|
+| delay | Delay in seconds for the metric evaluation | string | `600` | no |
+| environment | Architecture environment | string | - | yes |
+| failed_requests_threshold_critical | Maximum acceptable percent of failed requests | string | `5` | no |
+| filter_tags_custom | Tags used for custom filtering when filter_tags_use_defaults is false | string | `*` | no |
+| filter_tags_use_defaults | Use default filter tags convention | string | `true` | no |
+| message | Message sent when a Redis monitor is triggered | string | - | yes |
+| other_requests_threshold_critical | Maximum acceptable percent of other requests | string | `5` | no |
+| successful_requests_threshold_critical | Minimum acceptable percent of successful requests | string | `90` | no |
+| unauthorized_requests_threshold_critical | Maximum acceptable percent of unauthorized requests | string | `5` | no |
+
+Related documentation
+---------------------
+
+Azure API Management metrics documentation: [https://docs.microsoft.com/en-us/azure/api-management/api-management-howto-use-azure-monitor](https://docs.microsoft.com/en-us/azure/api-management/api-management-howto-use-azure-monitor)

cloud/azure/apimanagement/inputs.tf

@@ -0,0 +1,46 @@
+# Global Terraform
+variable "environment" {
+  description = "Architecture environment"
+  type = "string"
+}
+
+# Global DataDog
+variable "message" {
+  description = "Message sent when a Redis monitor is triggered"
+}
+
+variable "delay" {
+  description = "Delay in seconds for the metric evaluation"
+  default     = 600
+}
+
+variable "filter_tags_use_defaults" {
+  description = "Use default filter tags convention"
+  default     = "true"
+}
+
+variable "filter_tags_custom" {
+  description = "Tags used for custom filtering when filter_tags_use_defaults is false"
+  default     = "*"
+}
+
+# Azure API Management specific
+variable "failed_requests_threshold_critical" {
+  description = "Maximum acceptable percent of failed requests"
+  default = 5
+}
+
+variable "other_requests_threshold_critical" {
+  description = "Maximum acceptable percent of other requests"
+  default = 5
+}
+
+variable "unauthorized_requests_threshold_critical" {
+  description = "Maximum acceptable percent of unauthorized requests"
+  default = 5
+}
+
+variable "successful_requests_threshold_critical" {
+  description = "Minimum acceptable percent of successful requests"
+  default = 90
+}

cloud/azure/apimanagement/monitors-azure-apimanagement.tf

@@ -0,0 +1,160 @@
+data "template_file" "filter" {
+  template = "$${filter}"
+
+  vars {
+    filter = "${var.filter_tags_use_defaults == "true" ?
+             format("dd_monitoring:enabled,dd_azure_apimanagement:enabled,env:%s", var.environment) :
+             "${var.filter_tags_custom}"}"
+  }
+}
+
+resource "datadog_monitor" "apimgt_status" {
+  name    = "[${var.environment}] API Management status is not ok on {{name}}"
+  message = "${var.message}"
+
+  query = <<EOF
+      avg(last_5m):avg:azure.apimanagement_service.status{${data.template_file.filter.rendered}} by {name,resource_group,region} < 1
+      EOF
+  type  = "metric alert"
+
+  thresholds {
+    critical = 1
+  }
+
+  notify_no_data      = true
+  evaluation_delay    = "${var.delay}"
+  renotify_interval   = 0
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+  require_full_window = true
+  new_host_delay      = "${var.delay}"
+  no_data_timeframe   = 20
+
+  tags = ["env:${var.environment}", "resource:apimanagement", "team:azure", "provider:azure"]
+}
+
+resource "datadog_monitor" "apimgt_failed_requests" {
+  name    = "[${var.environment}] API Management {{name}} too much failed requests"
+  message = "${var.message}"
+
+  query = <<EOF
+    avg(last_5m): (
+      avg:azure.apimanagement_service.failed_requests{${data.template_file.filter.rendered}} * 100 /
+      avg:azure.apimanagement_service.total_requests{${data.template_file.filter.rendered}}
+      by {name,resource_group,region}
+    ) > ${var.failed_requests_threshold_critical}
+    EOF
+
+  thresholds {
+    critical  = "${var.failed_requests_threshold_critical}"
+  }
+
+  type                = "metric alert"
+  notify_no_data      = false
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+  require_full_window = true
+  new_host_delay      = "${var.delay}"
+  evaluation_delay    = "${var.delay}"
+  renotify_interval   = 0
+  no_data_timeframe   = 20
+
+  tags = ["env:${var.environment}", "resource:apimanagement", "team:azure", "provider:azure"]
+}
+
+resource "datadog_monitor" "apimgt_other_requests" {
+  name    = "[${var.environment}] API Management {{name}} too much other requests"
+  message = "${var.message}"
+
+  query = <<EOF
+    avg(last_5m): (
+      avg:azure.apimanagement_service.other_requests{${data.template_file.filter.rendered}} * 100 /
+      avg:azure.apimanagement_service.total_requests{${data.template_file.filter.rendered}}
+      by {name,resource_group,region}
+    ) > ${var.other_requests_threshold_critical}
+    EOF
+
+  thresholds {
+    critical  = "${var.other_requests_threshold_critical}"
+  }
+
+  type                = "metric alert"
+  notify_no_data      = false
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+  require_full_window = true
+  new_host_delay      = "${var.delay}"
+  evaluation_delay    = "${var.delay}"
+  renotify_interval   = 0
+  no_data_timeframe   = 20
+
+  tags = ["env:${var.environment}", "resource:apimanagement", "team:azure", "provider:azure"]
+}
+
+resource "datadog_monitor" "apimgt_unauthorized_requests" {
+  name    = "[${var.environment}] API Management {{name}} too much unauthorized requests"
+  message = "${var.message}"
+
+  query = <<EOF
+    avg(last_5m): (
+      avg:azure.apimanagement_service.failed_requests{${data.template_file.filter.rendered}} * 100 /
+      avg:azure.apimanagement_service.total_requests{${data.template_file.filter.rendered}}
+      by {name,resource_group,region}
+    ) > ${var.unauthorized_requests_threshold_critical}
+    EOF
+
+  thresholds {
+    critical  = "${var.unauthorized_requests_threshold_critical}"
+  }
+
+  type                = "metric alert"
+  notify_no_data      = false
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+  require_full_window = true
+  new_host_delay      = "${var.delay}"
+  evaluation_delay    = "${var.delay}"
+  renotify_interval   = 0
+  no_data_timeframe   = 20
+
+  tags = ["env:${var.environment}", "resource:apimanagement", "team:azure", "provider:azure"]
+}
+
+resource "datadog_monitor" "apimgt_successful_requests" {
+  name    = "[${var.environment}] API Management {{name}} successful requests rate too low"
+  message = "${var.message}"
+
+  query = <<EOF
+    avg(last_5m): (
+      avg:azure.apimanagement_service.successful_requests{${data.template_file.filter.rendered}} * 100 /
+      avg:azure.apimanagement_service.total_requests{${data.template_file.filter.rendered}}
+      by {name,resource_group,region}
+    ) < ${var.successful_requests_threshold_critical}
+    EOF
+
+  thresholds {
+    critical  = "${var.successful_requests_threshold_critical}"
+  }
+
+  type                = "metric alert"
+  notify_no_data      = true
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+  require_full_window = true
+  new_host_delay      = "${var.delay}"
+  evaluation_delay    = "${var.delay}"
+  renotify_interval   = 0
+  no_data_timeframe   = 20
+
+  tags = ["env:${var.environment}", "resource:apimanagement", "team:azure", "provider:azure"]
+}