Periskop: Exception Monitoring at Scale
summary
Describes Periskop, an open-source, pull-based exception monitoring system built at SoundCloud and inspired by Prometheus. It explains problems with push-based exception monitoring (thundering-herd on bad deployments, noisy logs), presents a pull model combined with aggregation and sampling to scale across many exceptions and instances, discusses limitations (short-lived processes, forked servers), and outlines Periskop’s implementation, client/server libraries, and planned features like a push-gateway, federation, time-series visualization and integrations.
AI-written summary · about
related talks
35mOpen Source Logging and Monitoring
36mObservability at Scale With Monitoring as Code...
29m“If you Don’t Monitor your Infrastructure, you Don’t Own it!” Regain Control Thanks to Prometheus
27mObservability with Prometheus and beyond
25mMonitoring in a Serverless World
35mMonitoring Microservices
34mMonitor My Socks: Using Prometheus in a Polyglot Open Source Microservices
40mThe Prometheus monitoring system