OS patching and upgrades with Amazon EC2 root volume replacement

This article explains how to use Amazon EC2 Replace Root Volume (RRV) to patch and upgrade operating systems on stateful workloads while preserving instance identity, IP addresses, and attached resources.

• Prepare and validate replacement root volumes on separate instances before swapping onto production
• Swap volumes with a single API call, limiting downtime to one reboot plus application restart time
• Preserve instance ID, private/public IP addresses, ENIs, data volumes, IAM profiles, and security groups during replacement
• Supports rollback by keeping original root volume or using snapshots as fallback
• For Auto Scaling groups, use Standby mode for one-at-a-time updates or instance refresh with ReplaceRootVolume strategy for fleet-wide updates
• Requirements include same Availability Zone, matching encryption, compatible boot mode, and ENA/NVMe drivers for Nitro instances

RRV reduces OS upgrade downtime and complexity by eliminating resource re-association steps required with traditional instance replacement.