Software has evolved tremendously over the past twenty years, but unfortunately the way that we both reason about and actually measure performance has barely changed. It’s time for people to stop thinking about software performance as a single number and see it as a shape.
In this talk, we’ll present the case for unsampled latency histograms as an evolution of and replacement for simple averages and percentile estimates.
- The historical context and evolution of in-application performance measurement, especially in the presence of microservices and FaaS/”serverless”.
- Why time-series statistics are not enough to explain the behavior of modern systems.
- How real-time latency histograms can radically improve the way we think about and manage application latency.
While p99 latency remains a useful statistic, the complexity of today’s microservice architectures requires a richer and more flexible approach. Our tools should identify, triage, and explain latency issues, especially as organizations adopt microservices and introduce multi-modal and highly-variable performance profiles as a consequence.
About the Speaker
Ben Sigelman is the cofounder and CEO of LightStep, where he and his colleagues deliver insights that put organizations back in control of their complex software applications. An expert in distributed tracing, Ben is the coauthor of the OpenTracing standard, a project within the Linux Foundation’s Cloud Native Computing Foundation (CNCF). Previously, he built Dapper, Google’s production distributed systems tracing infrastructure, and Monarch, Google’s fleet-wide time series collection, storage, analysis, and alerting system. Ben holds a BSc in mathematics and computer science from Brown University.