have its own metrics endpoint Hard API commitments within major versions No built-in TLS yet, use reverse proxies for now Richard Hartmann & Frederic Branczyk @TwitchiH & @fredbrancz Intro to Prometheus

security considerations Encryption, authentication, and authorization currently need to be handled via reverse proxies We will be changing Prometheus to support security out-of-the-box Richard Hartmann & Frederic

http_request_duration_seconds_bucket{job="n9e-proxy", le="0.1"} 500 http_request_duration_seconds_bucket{job="n9e-proxy", le="1"} 700 http_request_duration_seconds_bucket{job="n9e-proxy", le="10"} 850 http_request_ http_request_duration_seconds_bucket{job="n9e-proxy", le="20"} 1000 http_request_duration_seconds_bucket{job="n9e-proxy", le="+Inf"} 1000 根据这个数据，我们可以计算出落在各个延迟区间的请求数量，如下： 0 ~ 0.1 : 500 0.1 ~ 1 : 200 1 ~ 10 histogram_quantile(0.9, rate(http_request_duration_seconds_bucket[10m])) 上面的例子，是会对每个请求分别做计算，假设有两个模块：n9e-proxy、n9e-webapi，都统 计了 http_request_duration_seconds_bucket ，我们可能希望以模块为颗粒度，分别计算每 个模块的90分位延迟，写法是： histogram_quantile(