"How NOT to Measure Latency" by Gil Tene
9 人推荐 · 跨 9 个月 · 2015–2026
中文导读
Azul Systems 的 Gil Tene 深入剖析了测量延迟与响应时间时的经典误区，详细解释了为什么平均值会严重误导人，以及“协同遗漏”（coordinated omission）问题。他还介绍了 HdrHistogram 等开源工具，演示如何更严谨地度量和呈现性能数据。
为什么 HN 推它
社区讨论分位数、长尾延迟、负载测试或平均值陷阱时的必推神作，不少工程师直言这场分享彻底颠覆了自己对性能测试的认知。
「这是 Gil Tene 讲这个主题里真正能改变人认知的神级视频。」HN 网友原话：This a life changing video by Gil Tene on this topic.
英文原摘要（HackerTube / Claude 整理）
Gil Tene of Azul Systems explains common pitfalls in measuring latency and response time, including why averages mislead and the problem of coordinated omission. He introduces open source tools like HdrHistogram for better measurement and reporting.
The canonical recommendation whenever percentiles, tail latency, load testing or averages come up, described by some as life changing.
HN 上的提及
This is a summary of
none of this Coordinated Omission bullshit.
Gil Tene has an excellent talk[1] on this.
See also, Gil Tene
Gil Tene has an entire talk about this point called "How NOT to measure latency