system.schema.version version mismatch after failed Cassandra upgrade

Goal

Determine whether it is safe to manually align the system.schema.version table with the running Cassandra binary after an interrupted upgrade, or whether a pre‑upgrade snapshot must be restored before retrying the upgrade.

Constraints include verifying binary compatibility, avoiding schema corruption, and ensuring that subsequent nodetool upgradesstables processes all tables without skipping.

• What verification steps confirm that a manual UPDATE system.schema.version does not introduce schema inconsistencies?
• Under which conditions (e.g., mixed‑version clusters, specific release boundaries) is a snapshot restore preferable to a direct version‑table update?