Simulate a host maintenance event

Linux Windows

This page describes how to test the effects of your Compute Engine instance's host maintenance policy on your applications.

You might simulate a maintenance event on your VMs to test the following:

The effects of live migration on your applications.
How your applications and batch jobs handle preemption and shutdown when using one or more Spot VMs.
How your applications handle the shutdown and restart process for instances that are configured to terminate and restart during maintenance events rather than live migrate.
How workloads that are running on sole-tenant nodes behave during a host maintenance event, and see the effects of the sole-tenant VM's host maintenance policy on the applications running on the VMs.

If you try to simulate a host maintenance event on an instance that doesn't support live migration, the instance is either terminated or restarted, depending on the configured host maintenance policy.

Before you begin

Review the regional API rate limit for SimulateMaintenanceEventRequestsPerMinutePerProjectPerRegion.
If you haven't already, set up authentication. Authentication is the process by which your identity is verified for access to Google Cloud services and APIs. To run code or samples from a local development environment, you can authenticate to Compute Engine as follows.

Select the tab for how you plan to use the samples on this page:
gcloud
1. Install the Google Cloud CLI, then initialize it by running the following command:
  gcloud init
  Note: If you installed the gcloud CLI previously, make sure you have the latest version by running gcloud components update.
2. Set a default region and zone.
REST

To use the REST API samples on this page in a local development environment, you use the credentials you provide to the gcloud CLI.
For more information, see Authenticate for using REST in the Google Cloud authentication documentation.

Limitations

To correctly simulate a maintenance event on a sole-tenant node group that has a host maintenance policy set to migrate within node group, you need to trigger the maintenance event sequentially on each node.
For a sole-tenant node group, when you try to simulate a host maintenance event and the node group has a host maintenance policy set to migrate within the node group:
- If the number of nodes specified is less than or equal to the total number of holdback nodes that are reserved, then the host maintenance event simulation runs for all the specified nodes simultaneously.
- If the number of nodes specified is greater than the total number of reserved holdback nodes, then the simulation fails.
The number of maintenance event simulations you can start per minute per region is limited by the API rate limit for the simulate_maintenance_event_requests_per_region metric.

Simulate host maintenance events to test live migration

You can simulate a maintenance event for a compute instance by using either the Google Cloud CLI or an API request. This simulated event includes the different maintenance activities that occur in a regular maintenance event. This lets you observe the end-to-end process and test any automation that you might have implemented.

During the simulation of host maintenance event for an instance that uses live migration, the maintenance-event metadata key of the instance goes through the following changes:

At the start of the simulation, the value of the maintenance-event metadata key changes from NONE to MIGRATE_ON_HOST_MAINTENANCE.
Throughout the duration of the simulation event, the value remains as MIGRATE_ON_HOST_MAINTENANCE.
After the simulation ends, the value returns to NONE.

To query the maintenance event key, see Query the maintenance event metadata key.

gcloud

Use the compute instances simulate-maintenance-event command to simulate a maintenance event for an instance and test its configured host maintenance policy settings:

gcloud compute instances simulate-maintenance-event INSTANCE_NAME \
    --zone=ZONE --with-extended-notifications=True

Replace the following:

INSTANCE_NAME: the name of the compute instance where you want to simulate the maintenance event.

You can specify multiple instance names separated by single spaces to simulate maintenance events on more than one instance in the same zone. For example, instance-1 instance-2 instance-3.
ZONE: the zone where the instance is located.

REST

Construct a POST request to the compute.instances.simulateMaintenanceEvent method:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/ZONE/instances/INSTANCE_NAME/simulateMaintenanceEvent

Replace the following:

PROJECT_ID: the project ID for this request.
INSTANCE_NAME: the name of the instance for which you want to simulate the maintenance event.
ZONE: the zone where the instance is located.

Simulate host maintenance for compute instances that terminate

Additionally, by using the parameter --with-extended-notifications with a supported machine type you can test manually starting host maintenance during the simulated event.

gcloud

Use the compute instances simulate-maintenance-event command to simulate a maintenance event for an instance and test its configured host maintenance policy settings. You can optionally include the --with-extended-notifications flag.
```
gcloud compute instances simulate-maintenance-event INSTANCE_NAME \
   --zone=ZONE --with-extended-notifications=True
```
Replace the following:
- INSTANCE_NAME: the name of the compute instance where you want to simulate the maintenance event.
  
  You can specify multiple instance names separated by single spaces to simulate maintenance events on more than one instance in the same zone. For example, instance-1 instance-2 instance-3.
- ZONE: the zone where the instance is located.
Optional: To manually start the simulated maintenance event, use the compute instances perform-maintenance command.
```
gcloud compute instances perform-maintenance INSTANCE_NAME \
   --zone=ZONE
```
Replace the following:
- INSTANCE_NAME: the name of the compute instance where you want to simulate the maintenance event.
  
  You can specify multiple instance names separated by single spaces to simulate maintenance events on more than one instance in the same zone. For example, instance-1 instance-2 instance-3.
- ZONE: the zone where the instances are located.

REST

Construct a POST request to the compute.instances.simulateMaintenanceEvent method. You can optionally include the query parameter withExtendedNotifications.
```
POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/ZONE/instances/INSTANCE_NAME/simulateMaintenanceEvent?withExtendedNotifications=True
```
Replace the following:
- PROJECT_ID: the project ID for this request.
- INSTANCE_NAME: the name of the instance for which you want to simulate the maintenance event.
- ZONE: the zone where the instance is located.
Optional: To manually start the simulated maintenance event, construct a POST request to the compute.instances.performMaintenance method.
```
POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/ZONE/instances/INSTANCE_NAME/performMaintenance
```
Replace the following:
- INSTANCE_NAME: the name of the compute instance where you want to start the maintenance event.
  
  You can specify multiple instance names separated by single spaces to perform maintenance events on more than one instance in the same zone. For example, instance-1 instance-2 instance-3.
- ZONE: the zone where the instances are located.

Simulate host maintenance events on sole-tenant nodes

You can simulate a host maintenance event on sole-tenant nodes using either the Google Cloud CLI or an API request. During the simulation of the host maintenance event on a sole-tenant VM, the maintenance-event metadata key value doesn't change and remains NONE throughout the simulation.

gcloud

Run the sole-tenancy node-groups simulate-maintenance-event command to force sole-tenant nodes to activate their configured maintenance policy:

 gcloud compute sole-tenancy node-groups simulate-maintenance-event NODE_GROUP \
    --nodes=NODE_NAMES \
    --zone=ZONE \
    --async

Replace the following:

NODE_GROUP: the name of the node group where you want to simulate the maintenance event.
NODE_NAMES: the names of the nodes where you want to simulate the maintenance event. While specifying multiple node names, use comma-separated values, for example, node-1,node-2,node-3.
ZONE: the zone where the nodes are located.

REST

Construct a POST request to the compute.nodeGroups.simulateMaintenanceEvent method:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/ZONE/nodeGroups/NODE_GROUP/simulateMaintenanceEvent

{
  "nodes": [
      "NODE_NAMES"
  ]
}

Replace the following:

PROJECT_ID: the project ID for this request.
ZONE: the zone where the nodes are located.
NODE_GROUP: the name of the node group where you want to simulate the maintenance event.
NODE_NAMES: the names of the nodes where you want to simulate the maintenance event. Enclose the node name within double quotes, for example, "node-1". And, while specifying multiple node names, use comma-separated values, for example, "node-1","node-2","node-3".

What's next

Learn more about host events.
Learn more about host maintenance on sole-tenant nodes.
Learn more about setting VM host maintenance policy options.
Learn how to view VM operations.
Learn more about multi- and sole-tenancy.