Using Serverless Functions for Real-time Observability

Explore a conference talk from SREcon22 Americas that delves into extending a sub-second latency query engine for real-time observability using serverless functions. Learn how the system evolved from running in RAM to querying months of data using cloud storage and serverless functions. Discover the challenges of managing lambdas at scale, including issues with impatience, maximum concurrency, runtime and architecture configuration, and the price/performance considerations of utilizing 20,000 parallel workers. Gain insights from experienced speakers Liz Fong-Jones, a developer advocate and Site Reliability Engineer, and Jessica Kerr, a symmathecist focused on development teams as learning systems. Understand how this approach can enhance observability practices and improve system performance in large-scale environments.

SREcon22 Americas - Using Serverless Functions for Real-time Observability