Database Evolution: Partitioning and Indexing
summary
Argues that cloud-native databases become self-optimizing by combining adaptive partitioning (partition-level replication, geo-aware placement, dynamic shard splitting) with learned indexes and adaptive index advisors that trigger online rebuilds. Claims horizontal scaling adds capacity within minutes and that learned models replace B-trees for some workloads, reducing manual tuning. Uses Royal DB Bench and iTune (ML-driven config) to benchmark and quantify latency and operational-cost improvements, and stresses guardrails for safe automation.
AI-written summary · about
chapters (13)
- 0:00Welcome and Speaker Intro
- 1:01Why Databases Must Adapt
- 2:07Roadmap and Key Themes
- 2:44Adaptive Partitioning Basics
- 3:26Modern Partitioning Strategies
- 4:13Elastic Scaling and Availability
- 4:55Indexing Beyond B-Trees
- 5:25Adaptive and Learned Indexes
- 6:35Index Tradeoffs and Automation
- 7:14Machine Learning in Databases
- 8:53Benchmarks and Self-Optimizing Loop
- 9:41Governance and Implementation Path
- 10:32Wrap-Up and Q&A
related talks
27mScaling for Intelligence: Distributed SQL Databases in the Age of AI
30mThe database trends that are transforming your database infrastructure forever
44mRunning Non-Cloud-Native Databases in Cloud-Native Environments
44mContinuous Database Administration
51mDatabase systems: a decade of disruption and innovation
35mSmarter Databases in Kubernetes: Applying AI for Query Optimization and Self-Tuning
25mData-intensive PostgreSQL: Three ways to scale
22mOPTIMUSCLOUD: Heterogeneous Configuration Optimization for Distributed Databases