pk/atst
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

170237476 - Autoscaling fix

Browse Source 
Specifying the node count breaks the autoscaling min/max. When this
happens, the k8s cluster needs to be manually reconfigured. Terraform
does not remove the node count even when the node count option is
removed. The k8s cluster resource needed to be destroyed and re-created
in order to resolve the issue with node count and min/max options being
specified at the same time.
This commit is contained in:
Rob Gil 2019-12-30 12:57:57 -05:00
parent c14916dbe6
commit 0b7ff0679a
2 changed files with 1 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
terraform/modules/k8s/main.tf
View File

@ -19,7 +19,6 @@ resource "azurerm_kubernetes_cluster" "k8s" {
vm_size = "Standard_D1_v2"
os_disk_size_gb = 30
vnet_subnet_id = var.vnet_subnet_id
node_count = 1
enable_node_public_ip = true # Nodes need a public IP for external resources. FIXME: Switch to NAT Gateway if its available in our subscription
enable_auto_scaling = var.enable_auto_scaling
max_count = var.max_count # FIXME: if auto_scaling disabled, set to 0

2
terraform/providers/dev/k8s.tf
View File

@ -9,7 +9,7 @@ module "k8s" {
vnet_subnet_id = module.vpc.subnets #FIXME - output from module.vpc.subnets should be map
enable_auto_scaling = true
max_count = 5
min_count = 2
min_count = 3
}
#module "main_lb" {