nighthawk
nighthawk copied to clipboard
Optimise tracking of deferred timings in DelegatingRateLimiterImpl request-release timing
As of https://github.com/envoyproxy/nighthawk/pull/552, the DelegatingRateLimiter implementation buffers request-release timings from its inner rate limiter.
Today it uses a sorted std::list to accomplish this, but there are probably more efficient means to pull
this off when there are many timings to track.
Context: https://github.com/envoyproxy/nighthawk/pull/552/commits/131d2e04c141d8ff37ccb8a316d5639f828db1e1#r495628578