Mig Rolling Update Zero Downtime Capacity

What the solution is

A proactive rolling update in a Managed Instance Group (MIG) automatically rolls out a new instance template across existing instances according to the configured updatePolicy parameters, managing instance lifecycles while maintaining health and capacity.

How it meets the requirements

• Maintaining 100% capacity (zero drop below 30 instances): Setting maxUnavailable to 0 guarantees that Compute Engine will never stop, delete, or take offline any existing healthy VM instance before an equivalent replacement instance is fully booted, healthy, and available to serve traffic.
• Enabling new instance creation ahead of deletion: Setting maxSurge to a positive value (e.g., 3 instances or 10%) instructs the MIG Updater to provision temporary additional instances above the targetSize using the new instance template before terminating old instances.
• Supporting OS image updates: Updating an operating system boot disk image requires recreation of the VM; allowing the REPLACE action enables the MIG to replace instances with the new template without failing the update request.

Key advantages and capabilities

• Strict capacity guarantees: Eliminates under-provisioning risks by enforcing that total active capacity never dips below the configured target size.
• Controlled rollout pace: Allows engineers to scale maxSurge up or down to balance rollout speed against infrastructure quota and compute budget.
• Automated orchestration: The proactive updater manages instance creation, health check verification, and graceful decommission without manual intervention.

Why this is optimal

This configuration represents the standard pattern for zero-downtime rolling upgrades in production environments where capacity deficits cannot be tolerated. The Updater provisions new surged VMs first, verifies their health, and only then tears down old instances.