Update AKS logging to log fewer things, add flexibility in disk config for system pool (#94)

Author: lumberbaron

aks/terraform/main.tf

@@ -66,6 +66,7 @@ module "cluster" {
   outbound_ip_count        = var.outbound_ip_count
   outbound_ports_allocated = var.outbound_ports_allocated
 
+  worker_node_vm_size        = local.system_vm_size
   worker_node_ssh_public_key = var.worker_node_ssh_public_key
 
   kubernetes_api_public_access       = var.kubernetes_api_public_access
@@ -83,6 +84,8 @@ module "cluster" {
 locals {
   os_disk_size_gb = 48
 
+  system_vm_size = "Standard_D2s_v3"
+
   prod1k_vm_size     = "Standard_E2s_v3"
   prod10k_vm_size    = "Standard_E4s_v3"
   prod100k_vm_size   = "Standard_E8s_v3"

aks/terraform/modules/broker-node-pool/README.md

@@ -37,6 +37,7 @@ No modules.
 | <a name="input_node_pool_taints"></a> [node\_pool\_taints](#input\_node\_pool\_taints) | Kubernetes taints added to worker nodes in the node pools. | `list(string)` | n/a | yes |
 | <a name="input_subnet_id"></a> [subnet\_id](#input\_subnet\_id) | The subnet that will contain the worker nodes in each node pool. | `string` | n/a | yes |
 | <a name="input_worker_node_disk_size"></a> [worker\_node\_disk\_size](#input\_worker\_node\_disk\_size) | The OS disk size (in GB) used for the worker nodes in each node pool. | `string` | n/a | yes |
+| <a name="input_worker_node_disk_type"></a> [worker\_node\_disk\_type](#input\_worker\_node\_disk\_type) | The type of the OS disk for the worker nodes in each node pool. | `string` | `"Ephemeral"` | no |
 | <a name="input_worker_node_vm_size"></a> [worker\_node\_vm\_size](#input\_worker\_node\_vm\_size) | The VM size used for the worker nodes in each node pool. | `string` | n/a | yes |
 
 ## Outputs

aks/terraform/modules/broker-node-pool/main.tf

@@ -22,7 +22,7 @@ resource "azurerm_kubernetes_cluster_node_pool" "this" {
   vnet_subnet_id = var.subnet_id
 
   vm_size         = var.worker_node_vm_size
-  os_disk_type    = "Ephemeral"
+  os_disk_type    = var.worker_node_disk_type
   os_disk_size_gb = var.worker_node_disk_size
 
   node_labels = var.node_pool_labels

aks/terraform/modules/broker-node-pool/variables.tf

@@ -40,6 +40,12 @@ variable "worker_node_disk_size" {
   description = "The OS disk size (in GB) used for the worker nodes in each node pool."
 }
 
+variable "worker_node_disk_type" {
+  type        = string
+  default     = "Ephemeral"
+  description = "The type of the OS disk for the worker nodes in each node pool."
+}
+
 variable "node_pool_labels" {
   type        = map(string)
   description = "Kubernetes labels added to worker nodes in the node pools."

aks/terraform/modules/cluster/README.md

@@ -54,7 +54,10 @@ No modules.
 | <a name="input_resource_group_name"></a> [resource\_group\_name](#input\_resource\_group\_name) | The name of the resource group that will contain the cluster. | `string` | n/a | yes |
 | <a name="input_route_table_id"></a> [route\_table\_id](#input\_route\_table\_id) | The ID of the route table of the subnet where the cluster will reside. | `string` | n/a | yes |
 | <a name="input_subnet_id"></a> [subnet\_id](#input\_subnet\_id) | The ID of the subnet where the cluster will reside. | `string` | n/a | yes |
+| <a name="input_worker_node_os_disk_size_gb"></a> [worker\_node\_os\_disk\_size\_gb](#input\_worker\_node\_os\_disk\_size\_gb) | The size of the OS disk for the worker nodes in the default (system) node pool. | `number` | `48` | no |
+| <a name="input_worker_node_os_disk_type"></a> [worker\_node\_os\_disk\_type](#input\_worker\_node\_os\_disk\_type) | The type of the OS disk for the worker nodes in the default (system) node pool. | `string` | `"Ephemeral"` | no |
 | <a name="input_worker_node_ssh_public_key"></a> [worker\_node\_ssh\_public\_key](#input\_worker\_node\_ssh\_public\_key) | The public key that will be added to the authorized keys file on the worker nodes for SSH access. | `string` | n/a | yes |
+| <a name="input_worker_node_vm_size"></a> [worker\_node\_vm\_size](#input\_worker\_node\_vm\_size) | The default VM size for the worker nodes in the default (system) node pool. | `string` | n/a | yes |
 
 ## Outputs
 

aks/terraform/modules/cluster/main.tf

@@ -1,8 +1,5 @@
 locals {
   worker_node_username = "worker"
-
-  os_disk_size_gb = 48
-  default_vm_size = "Standard_D2s_v3"
 }
 
 ################################################################################
@@ -62,9 +59,9 @@ resource "azurerm_kubernetes_cluster" "cluster" {
   default_node_pool {
     name            = "default"
     node_count      = 2
-    vm_size         = local.default_vm_size
-    os_disk_size_gb = local.os_disk_size_gb
-    os_disk_type    = "Ephemeral"
+    vm_size         = var.worker_node_vm_size
+    os_disk_size_gb = var.worker_node_os_disk_size_gb
+    os_disk_type    = var.worker_node_os_disk_type
     vnet_subnet_id  = var.subnet_id
     zones           = var.availability_zones
     max_pods        = var.max_pods_per_node
@@ -156,15 +153,12 @@ resource "azurerm_monitor_diagnostic_setting" "cluster" {
     category = "cluster-autoscaler"
   }
 
-  enabled_log {
-    category = "kube-apiserver"
-  }
-
   enabled_log {
     category = "kube-audit-admin"
   }
 
-  enabled_log {
-    category = "kube-controller-manager"
+  metric {
+    category = "AllMetrics"
+    enabled  = false
   }
 }

aks/terraform/modules/cluster/variables.tf

@@ -109,4 +109,21 @@ variable "max_pods_per_node" {
   type        = number
   default     = 110
   description = "The maximum number of pods for the worker nodes in the node pools."
+}
+
+variable "worker_node_vm_size" {
+  type        = string
+  description = "The default VM size for the worker nodes in the default (system) node pool."
+}
+
+variable "worker_node_os_disk_size_gb" {
+  type        = number
+  default     = 48
+  description = "The size of the OS disk for the worker nodes in the default (system) node pool."
+}
+
+variable "worker_node_os_disk_type" {
+  type        = string
+  default     = "Ephemeral"
+  description = "The type of the OS disk for the worker nodes in the default (system) node pool."
 }