Ticketing is a distributed lock server built to handle your highest-traffic moments as fast as possible. It exists to stop the accidents that happen when a limited quantity oversells, the same seat gets assigned to two people, or a single coupon gets redeemed twice at once.

Unlike most distributed locks, which are implemented as scripts, Ticketing is written in Rust to get both memory efficiency and speed. It reads the fixed-width binary protocol by offset with no parsing, and doesn't allocate on the heap even while handling a request. In single-node benchmarks it handles over 1 million ops/sec while using under 10MB of memory. Its queue guarantees ordering, and with the Raft cluster turned on, service keeps running with zero downtime even as nodes go down one at a time — say, during cloud instance reboots or a rolling maintenance pass.