Under the hood · provably fair

A global crowd agrees on one falling price — and everyone settles fairly.

Provable, not promised. Every claim on this page is a live query from Aurora DSQL — two active-active regions, strong consistency.

Live · Tokyo (ap-northeast-1) + Seoul (ap-northeast-2) · refreshes every 3 s

Everyone paid one price

Settling a proof rally — this updates automatically.

One price, two regions

Loading region data…

Proven at scale

Real numbers from a load test on Tokyo Aurora DSQL — not simulated, not claimed. The correctness properties are identical at any N.

2,500

join attempts

398

OCC retries → success

$100→$52

price melt

1

distinct price paid

$0.00

ledger drift

0

double-joins / charges

3.6 s

settle 2,000 users

Why it needs this stack

Live price is a derived read

tier(sum(count_shards))

The price is computed from the authoritative count on every read — never stored, never raced. Crossing a tier threshold lowers the price for everyone atomically. No stale tier value; no transition race to lose.

128-shard counter per rally

random_shard() → count++

A single count++ row is the classic serializable-database contention killer. Random shard selection spreads the hot write across 128 rows, keeping OCC conflicts rare and all retrying cleanly to success.

Idempotent joins by design

INSERT … ON CONFLICT DO NOTHING

A unique (rally_id, user_id) constraint means a double-tap, a browser retry, or a race between two windows never double-counts. The ledger only charges you once — enforced at the database level.

DynamoDB handles the fan-out

liveFrame → DDB → SSE

The live frame is projected to DynamoDB on every committed join. Watchers subscribe via SSE — never via DSQL aggregate scans. Correctness lives in DSQL; million-scale fan-out lives in DynamoDB.