System Design

The guiding question

How do load balancers, caches, databases and queues fit together into a system that survives real traffic and keeps running when parts break?

Episodes

1. 1Client and Server: The Journey of a Request (HTTP, DNS)Published
2. 2API Design: REST and RPCComing soon
3. 3Load Balancing: Spreading Traffic Over Many MachinesComing soon
4. 4Caching and CDNs: Hit Rate Versus FreshnessComing soon
5. 5Database Basics: SQL and NoSQLComing soon
6. 6Message Queues: Smoothing Peaks and DecouplingComing soon
7. 7Security Basics: Authentication, Authorization and TLSComing soon
8. 8Functional and Non-Functional RequirementsComing soon
9. 9Capacity Estimation: Back-of-the-Envelope MathComing soon
10. 10Latency Numbers: A Ruler Drawn to ScaleComing soon
11. 11Data Modeling: Normalized or DenormalizedComing soon
12. 12Sharding: Shard Keys and Hot SpotsComing soon
13. 13Replication by Design: Lag, Failover and Lost WritesComing soon
14. 14Transactions and Isolation LevelsComing soon
15. 15Search and the Inverted IndexComing soon
16. 16Availability and the NinesComing soon
17. 17Rate Limiting: Token Bucket and Leaky BucketComing soon
18. 18Levels of ConsistencyComing soon
19. 19Failure and Recovery: RPO and RTOComing soon
20. 20Observability: Logs, Metrics and TracesComing soon
21. 21Case Study: A URL ShortenerComing soon
22. 22Case Study: Designing a Distributed Rate LimiterComing soon
23. 23Case Study: A News FeedComing soon
24. 24Case Study: A Chat SystemComing soon
25. 25Case Study: Cloud Storage and Multi-Device SyncComing soon
26. 26Case Study: A Video PlatformComing soon
27. 27Case Study: Search AutocompleteComing soon
28. 28Case Study: Nearby FriendsComing soon
29. 29Capstone: Assembling the Parts Into a Whole SystemComing soon