Talk abstract
Why Shipping Native Histograms To Our Users Is a Game Changer
Traditional histograms in Prometheus have their shortcomings. Some of them are much more significant when you are building and shipping software than when you are and end user. Let's take deciding on bucket distribution for a histogram as an example. While a certain distribution might make sense in a local Kubernetes cluster, it might not make sense at scale, with 100s of clusters or nodes. Or how about being tired of losing visibility to the 100th percentile of your latency? Join Krisztian and learn about the benefits of using Native Histograms trough real-life examples with solo.io's Gloo products!