Hedged request.
also: hedged requests · request hedging · backup request · backup requestsIf a request hasn't answered by about the P95 time, send a duplicate to another server and use whichever replies first.
the two-minute refresher
- Send to server A; wait the P95 time; if nothing, also send to B; take the first answer and cancel the other.
- It cuts tail latency for a few percent extra load.
- Only for idempotent (safe to repeat) requests.