How would you upgrade a subnet's VM without downtime?

• A. Upgrade all at once without testing.
• B. Plan a staged VM upgrade, test in staging, coordinate via governance, and perform rolling upgrades with compatibility checks.
• C. Disable governance during upgrade.
• D. Ignore compatibility checks.

Answer: (B)

Upgrading a subnet’s VM without downtime comes from doing a careful, staged rollout rather than a big, all-at-once change. The main idea is to test first, coordinate with the community, and roll out the upgrade in small, verifiable steps while checking that everything stays compatible at each stage.

Start by planning and testing in a staging environment that mirrors production. This lets you catch issues with the new VM version without affecting real users or the live network. Once testing looks good, bring in governance to review and approve the upgrade plan. Governance ensures all stakeholders are aware, methods are safe, and there’s a clear path to rollback if something goes wrong.

Then execute the upgrade in a rolling fashion. upgrade a subset of VM instances or validators first, monitor performance, correctness, and compatibility, and only proceed to the next group once the first one is confirmed healthy. This staggered approach means a problem in one part doesn’t knock out the entire subnet, and you can halt or revert that subset if needed while leaving the rest online.

Compatibility checks are essential throughout. They verify that the new VM version speaks the same protocol expectations, handles existing state correctly, and interoperates with other components without breaking consensus or transaction processing. This careful verification prevents subtle incompatibilities from causing downtime later.

Why not upgrade everything at once or skip safeguards? A single, rushed upgrade can expose the whole network to unforeseen failures and lengthy downtime if something goes wrong. Dismissing governance removes critical oversight and risk controls. Ignoring compatibility checks invites hard-to-detect breakages that can undermine the network’s reliability. The staged, tested, and governed rolling upgrade with compatibility checks is what makes a downtime-free upgrade feasible.