After syncing to head with --syncmode snap, stopping the node, and restarting with --syncmode full, the node still runs in snap sync. Logs show snap-specific activity (e.g., “Inserting snap-sync blocks”) and never log a successful switch to full sync.