Vostok Hercules: 3 Years Delivering Telemetry — Flight Normal
summary
An examination of a high-throughput telemetry processing and delivery system, Vostok Hercules, built around Apache Kafka. Covers practical pros and cons of an event-driven architecture used in production for three years, with concrete experience handling telemetry data at scale.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 3:28Scalable architecture challenges
- 6:59What is telemetry
- 9:24Hercules system architecture
- 12:13Event-driven architecture
- 16:29Async and consistency
- 18:12Real-time trace analytics
- 22:26Data-driven load balancing
- 25:20Distributed systems issues
- 33:04Duplicate events problem
- 43:11Choosing Kafka as foundation
- 51:14Lambda architecture and Kafka
- 55:49Q&A
related talks
49mThinking in time series, the new solution to your pro...
36mTelemetry Analysis for Scaling
58mAstarte: from data collection to fleet management
36mTelemetry Analysis at Scale
44mProcess and analyze millions of events per second with open source tools
35mReal-time analytics at IoT scale
47mBuilding High Performance Real-time Stream Processing Microservices
30mData Platform Infrastructure at Rivian: A Deep Dive