Docs / Operators / Upgrades

upgrade without guessing.

This guide defines the operator workflow for validating releases, scheduling upgrades, protecting state and recovering cleanly if a rollout fails.

before upgrading

Confirm the target release, compatibility window, expected migration behavior and any network activation height before changing a production node. Verify release artifacts and preserve a known-good rollback path.

maintenance planning

Coordinate validator availability, RPC capacity and indexer catch-up so an upgrade does not create avoidable downtime or data gaps. Keep enough healthy capacity online during staggered rollouts.

state protection

Take verified backups or snapshots as appropriate for the service. Record the currently running version, configuration checksum and relevant chain state so rollback decisions are based on known facts.

post-upgrade checks

Confirm process health, peer connectivity, chain progress, signing behavior, RPC correctness, index freshness and alert state before considering the upgrade complete.

rollback

If the new release cannot safely remain in service, stop cleanly, restore the supported prior state or binary path and revalidate synchronization before returning the node to normal operation.

release-specific values

Exact binaries, checksums, activation heights, migration commands and compatibility notes will be published only after each release is qualified.