From 2a4bf3956934d236eacf867ff0a21ee460913e7e Mon Sep 17 00:00:00 2001
From: Quentin Manfroi <quentin.manfroi@fr.clara.net>
Date: Thu, 11 Jul 2019 18:45:51 +0200
Subject: [PATCH] MON-455 add terminated container monitor

---
 caas/kubernetes/pod/README.md           |  9 ++++++
 caas/kubernetes/pod/inputs.tf           | 40 +++++++++++++++++++++++++
 caas/kubernetes/pod/monitors-k8s-pod.tf | 30 +++++++++++++++++++
 caas/kubernetes/pod/outputs.tf          |  5 ++++
 4 files changed, 84 insertions(+)

diff --git a/caas/kubernetes/pod/README.md b/caas/kubernetes/pod/README.md
index 1dbae32..c6dd2b5 100644
--- a/caas/kubernetes/pod/README.md
+++ b/caas/kubernetes/pod/README.md
@@ -17,6 +17,7 @@ module "datadog-monitors-caas-kubernetes-pod" {
 Creates DataDog monitors with the following checks:
 
 - Kubernetes Pod phase status failed
+- Kubernetes Pod terminated anormaly
 - Kubernetes Pod waiting errors
 
 ## Inputs
@@ -43,6 +44,13 @@ Creates DataDog monitors with the following checks:
 | pod\_phase\_status\_time\_aggregator | Monitor aggregator for Pod phase status [available values: min, max or avg] | string | `"max"` | no |
 | pod\_phase\_status\_timeframe | Monitor timeframe for Pod phase status [available values: `last_#m` (1, 5, 10, 15, or 30), `last_#h` (1, 2, or 4), or `last_1d`] | string | `"last_5m"` | no |
 | prefix\_slug | Prefix string to prepend between brackets on every monitors names | string | `""` | no |
+| terminated\_enabled | Flag to enable Pod terminated monitor | string | `"true"` | no |
+| terminated\_extra\_tags | Extra tags for Pod terminated monitor | list(string) | `[]` | no |
+| terminated\_message | Custom message for Pod terminated monitor | string | `""` | no |
+| terminated\_threshold\_critical | terminated critical threshold | string | `"0.5"` | no |
+| terminated\_threshold\_warning | terminated warning threshold | string | `"0"` | no |
+| terminated\_time\_aggregator | Monitor aggregator for Pod terminated [available values: min, max or avg] | string | `"sum"` | no |
+| terminated\_timeframe | Monitor timeframe for Pod terminated [available values: `last_#m` (1, 5, 10, 15, or 30), `last_#h` (1, 2, or 4), or `last_1d`] | string | `"last_10m"` | no |
 
 ## Outputs
 
@@ -50,6 +58,7 @@ Creates DataDog monitors with the following checks:
 |------|-------------|
 | error\_id | id for monitor error |
 | pod\_phase\_status\_id | id for monitor pod_phase_status |
+| terminated\_id | id for monitor terminated |
 
 ## Related documentation
 
diff --git a/caas/kubernetes/pod/inputs.tf b/caas/kubernetes/pod/inputs.tf
index 73f74da..b78b617 100644
--- a/caas/kubernetes/pod/inputs.tf
+++ b/caas/kubernetes/pod/inputs.tf
@@ -110,3 +110,43 @@ variable "error_threshold_warning" {
   description = "error warning threshold"
 }
 
+variable "terminated_enabled" {
+  description = "Flag to enable Pod terminated monitor"
+  type        = string
+  default     = "true"
+}
+
+variable "terminated_extra_tags" {
+  description = "Extra tags for Pod terminated monitor"
+  type        = list(string)
+  default     = []
+}
+
+variable "terminated_message" {
+  description = "Custom message for Pod terminated monitor"
+  type        = string
+  default     = ""
+}
+
+variable "terminated_time_aggregator" {
+  description = "Monitor aggregator for Pod terminated [available values: min, max or avg]"
+  type        = string
+  default     = "sum"
+}
+
+variable "terminated_timeframe" {
+  description = "Monitor timeframe for Pod terminated [available values: `last_#m` (1, 5, 10, 15, or 30), `last_#h` (1, 2, or 4), or `last_1d`]"
+  type        = string
+  default     = "last_10m"
+}
+
+variable "terminated_threshold_critical" {
+  default     = 0.5
+  description = "terminated critical threshold"
+}
+
+variable "terminated_threshold_warning" {
+  default     = 0
+  description = "terminated warning threshold"
+}
+
diff --git a/caas/kubernetes/pod/monitors-k8s-pod.tf b/caas/kubernetes/pod/monitors-k8s-pod.tf
index ab5154c..0d66011 100644
--- a/caas/kubernetes/pod/monitors-k8s-pod.tf
+++ b/caas/kubernetes/pod/monitors-k8s-pod.tf
@@ -57,3 +57,33 @@ EOQ
   tags = concat(["env:${var.environment}", "type:caas", "provider:kubernetes", "resource:kubernetes-pod", "team:claranet", "created-by:terraform"], var.error_extra_tags)
 }
 
+resource "datadog_monitor" "terminated" {
+  count   = var.terminated_enabled == "true" ? 1 : 0
+  name    = "${var.prefix_slug == "" ? "" : "[${var.prefix_slug}]"}[${var.environment}] Kubernetes Pod terminated anormaly"
+  message = coalesce(var.terminated_message, var.message)
+  type    = "query alert"
+
+  query = <<EOQ
+    ${var.terminated_time_aggregator}(${var.terminated_timeframe}):
+      sum:kubernetes_state.container.status_report.count.terminated${module.filter-tags-nocontainercreating.query_alert} by {namespace,pod,reason}
+    > ${var.terminated_threshold_critical}
+EOQ
+
+  thresholds = {
+    critical = var.terminated_threshold_critical
+    warning  = var.terminated_threshold_warning
+  }
+
+  evaluation_delay    = var.evaluation_delay
+  new_host_delay      = var.new_host_delay
+  notify_no_data      = false
+  renotify_interval   = 0
+  notify_audit        = false
+  timeout_h           = 0
+  include_tags        = true
+  locked              = false
+  require_full_window = true
+
+  tags = concat(["env:${var.environment}", "type:caas", "provider:kubernetes", "resource:kubernetes-pod", "team:claranet", "created-by:terraform"], var.terminated_extra_tags)
+}
+
diff --git a/caas/kubernetes/pod/outputs.tf b/caas/kubernetes/pod/outputs.tf
index 70cf2f9..69f066e 100644
--- a/caas/kubernetes/pod/outputs.tf
+++ b/caas/kubernetes/pod/outputs.tf
@@ -8,3 +8,8 @@ output "error_id" {
   value       = datadog_monitor.error.*.id
 }
 
+output "terminated_id" {
+  description = "id for monitor terminated"
+  value       = datadog_monitor.terminated.*.id
+}
+