Maintenance policy management

Create a VM with maintenance policies:

FOLDER_ID=$(yc config get folder-id)

yc compute instance create --name mnt-vm1 --hostname mnt-vm1\
  --folder-id $FOLDER_ID \
  --zone ru-central1-a \
  --create-boot-disk image-folder-id=standard-images,image-family=ubuntu-2204-lts \
  --cores=2 --memory=4G --core-fraction=100 \
  --network-interface subnet-name=default-ru-central1-a,ipv4-address=auto,nat-ip-version=ipv4 \
  --ssh-key ~/.ssh/id_ed25519.pub \
  --maintenance-policy restart \
  --maintenance-grace-period 1m

1. Create a Terraform manifest named mnt-vm.tf:

   terraform {
     required_providers {
       yandex = {
         source  = "yandex-cloud/yandex"
         version = "~> 0.105.0"
       }
     }
   }
   
   variable "folder_id" {
     type = string
   }
   
   data "yandex_compute_image" "vm_image" {
     family = "ubuntu-2204-lts"
   }
   
   data "yandex_vpc_subnet" "vm_subnet" {
     folder_id = var.folder_id
     name      = "default-ru-central1-a"
   }
   
   resource "yandex_compute_instance" "vm_instance" {
     name                     = "mnt-vm1"
     folder_id                = var.folder_id
     hostname                 = "mnt-vm1"
     zone                     = "ru-central1-a"
     maintenance_policy       = "restart" # restart| migrate | unspecified
     maintenance_grace_period = "1m"
   
     resources {
       cores  = 2
       memory = 4
     }
   
     boot_disk {
       initialize_params {
         image_id = data.yandex_compute_image.vm_image.id
       }
     }
   
     network_interface {
       subnet_id = data.yandex_vpc_subnet.vm_subnet.id
       nat       = true
     }
   
     metadata = {
       ssh-keys = "ubuntu:${file("~/.ssh/id_ed25519.pub")}"
     }
   }
   

2. Apply the Terraform manifest:

   export YC_TOKEN=$(yc iam create-token)
   export TF_VAR_folder_id=$(yc config get folder-id)
   
   terraform init
   terraform apply
   

Use the create REST API method for the Instance resource or the InstanceService/Create gRPC API call.

When creating the VM, specify the maintenance policy settings: RESTART as the policy type (maintenancePolicy) and 60 seconds as the waiting period (maintenanceGracePeriod).

You can change the VM maintenance policy type by specifying a new value for the --maintenance-policy parameter through the following command:

yc compute instance update --name=mnt-vm1 --maintenance-policy=migrate