Schema design for fast applications
summary
Shows concrete MongoDB patterns and numbers: reproducing a TPC-C style order workload reduces data and index size and raises throughput by about 15–25% versus relational joins using a TPC-C benchmark and MongoDB with WiredTiger block manager explained. High write OLTP favors many small documents (Amazon found ~90% writes) while read hotspots benefit from limited embedding like the computed pattern and last‑10 comments subset. Changing a trade model removed a needless index and boosts performance 2–3×, so the schema choice determines cost and latency.
AI-written summary · about
chapters (11)
- 0:00Introduction and performance
- 2:32Dimensions of performance
- 5:22Relational vs NoSQL modeling
- 7:33Understanding document storage
- 12:11Avoiding anti-patterns
- 15:04Designing for access patterns
- 18:32Applying modeling techniques
- 25:27Optimizing order workflows
- 33:41Schema validation importance
- 36:50Real-world benchmarks
- 40:19Trading and final thoughts
related talks
36mMongoDB Schema Design - best practices
44mMongoDB: optimize your indexing
53mFrom RDBMS to NoSQL (PRT314)
01h 00mThe Modern Database Debate: Documents, Tables, or Documents in Tables?
44mThe Developer's Guide to Data Modelling with Document Databases
53mNoSQL Best Practices for PostgreSQL
26mNoSQL Data Modeling for Front-end Developers
39mMongoDB vs MongoDB: How, When and Where to Use It