Booking.com employs a unique approach to observability by using events instead of the traditional logs, metrics, and traces. Events provide a fuller context and can generate classical observability pillars while enabling advanced analytics. Booking's observability system includes an event-proxy daemon and utilizes Kafka to process event data, transforming it into traces, application performance metrics, and error logs for debugging.
Table of contents
The three pillars of observabilityWhat are events, and why does Booking use themHow Booking implemented eventsReferencesFood for thoughtsInteresting ReadsSort: