Book Overflow Reliability, Scalability, and Maintainability - Designing Data-Intensive Applications by Martin Kleppman
Jan 19, 2026
Dive into the intricacies of designing data-intensive applications with insights on reliability, scalability, and maintainability. Explore the Twitter scalability challenge and how chaos engineering fosters resilience in systems. Discover the implications of AI-generated code and its effects on the code review process. Learn about data models, query languages, and the trade-offs between various database strategies. Finally, get practical takeaways that highlight the timelessness of these concepts for engineers.
AI Snips
Chapters
Books
Transcript
Episode notes
Design For Fault Tolerance
- Faults do not equal failures; design systems to tolerate faults to avoid user-facing failures.
- Carter Morgan emphasizes chaos engineering and planning for hardware, software, and human faults.
Invert Read-Heavy Workloads
- Invert heavy read workflows: precompute or push updates for common reads rather than querying many followers on demand.
- Use hybrid approaches for extreme cases like celebrity accounts to avoid massive fan-out writes.
Measure Tail Latency Not Just Averages
- Percentiles reveal tail latency problems that averages hide and matter for user experience.
- Tail requests (P99.9) often correspond to your most valuable customers and deserve attention.
