maxPendingLatency

Maximum amount of time that a request should wait in the pending queue before starting a new instance to handle it. A duration in seconds with up to nine fractional digits, terminated by 's'. Example: "3.5s".