Public Preview: Automatic Zone Placement for Virtual Machine Scale Sets

We’re thrilled to share the public preview of Automatic Zone Placement for Azure Virtual Machine Scale Sets. This exciting new feature allows Azure to automatically choose availability zones for your scale set, taking into account SKU availability, capacity, and your specific placement preferences.

With Automatic Zone Placement, zonal deployments become much simpler. You won’t need to monitor SKU availability or manage fixed zone lists for each Azure region anymore. This feature works seamlessly with scale sets, whether you’re using Uniform or Flexible orchestration modes.

Streamlining Zonal Deployments Across Regions

When setting up a zonal scale set, you usually need to indicate which availability zones to use for your instances. However, SKU availability and capacity can fluctuate between zones and regions. Therefore, a setup that works in one location might call for a different zone in another, which can complicate large-scale and multi-region deployments.

Thanks to Automatic Zone Placement, you can simply define your placement requirements, and Azure will take care of selecting the zones that best align with your needs.

Benefits of Automatic Zone Placement

• Consistent Configuration Across Regions: You can apply the same scale set configuration in different regions without needing to maintain specific zone lists for each one.
• Capacity-Aware Deployments: Azure considers the current capacity when picking availability zones for your deployment.
• Adjustable Resiliency Controls: You have the freedom to set how many zones the scale set may utilise and the maximum capacity allowed in each zone.
• Enhanced Allocation Flexibility: The scale set can tap into additional eligible zones if its current zones can’t meet a scale-out request.
• Works with New and Existing Scale Sets: You can enable Automatic Zone Placement while creating a scale set or shift an existing one from zones you’ve chosen.

How Automatic Zone Placement Functions

When you activate Automatic Zone Placement, Azure analyses SKU availability, capacity, the current zone distribution of the scale set, and any placement rules you’ve set before positioning new VM instances.

If you enable Automatic Zone Placement but don’t specify additional settings, Azure follows its default guidelines:

• The goal is an even distribution across three availability zones, with a minimum spread across at least two zones.
• Limits any zone to a maximum of 50% of the total instances in the scale set.
• Can utilise additional availability zones (3+) if the current zones can’t accommodate a scale-out request and another zone is available.

You can customise placement settings as needed to optimise your workload. This includes defining eligible zones, setting a maximum number of zones, determining the maximum percentage of instances that can be allocated to one zone, and adjusting balancing behaviour. This ensures you can find the perfect balance between successful allocation and zonal distribution for your workload.

Do note that Automatic Zone Placement only applies when Azure creates new instances. It won’t affect existing VMs, but if you need to address imbalances among existing instances, consider using Automatic Zone Balance.

How to Get Started

You can configure Automatic Zone Placement via the Azure portal, Azure CLI, Azure PowerShell, or the REST API.

To start using this feature during the public preview, follow these steps:

1. Register the “Microsoft.Compute/VmssAutomaticZonePlacement” feature for your subscription.
2. Ensure you are using Compute API version 2026-03-01 or later.
3. Create or update a scale set in a public Azure region that supports availability zones.
4. Set the zone placement policy to Auto and decide on any optional placement constraints.

For details on prerequisites, limitations, and examples of how to configure this for different deployment methods, check out Configure automatic zone placement for a scale set.