Build & scale applications with Azure SQL Hyperscale
summary
Hyperscale separates compute and storage and scales storage automatically to 100 TB storage, exposes provisioned compute from 2–128 vCores, and offers a serverless preview that autos-scales CPU/memory with an 8x min→max ratio. Read scale uses named replicas (up to 30 replicas) and a demo sustains about 26,000 batch requests/sec on a read replica while combined workloads hit ~40–45k, cross-region failover completes in roughly 30 second cross-region failover, and premium CPUs and memory‑optimized SKUs raise throughput significantly.
AI-written summary · about
chapters (14)
- 0:00Introduction to Hyperscale
- 2:25Scalability and compute options
- 5:38Hyperscale architecture deep dive
- 13:05Read operations and caching
- 14:51Named replicas for read scaling
- 17:22Geo-replication and failover
- 20:55Demo of replica topology
- 26:29Serverless in Azure SQL
- 30:45Hyperscale serverless demo
- 35:38Premium series and performance
- 37:47Premium series demo
- 41:24Memory optimized premium series
- 44:58Backup retention updates
- 45:40Q&A session
related talks
51mAzure SQL Hyperscale Deep Dive
17mAutomatic compute scaling with serverless for Hyperscale in Azure SQL Databases
52mMore for less: Cost optimizing your Azure SQL databases
51mFly at Mach-speed with Azure SQL Managed Instance
01h 11mHand-On-Labs: How crazy are these from Azure Database PaaS!
47mPerfecting business continuity for Azure SQL DB
01h 15mChoose Azure SQL Database and Learn What Every Developer Should Know
49mSQL Server on Azure VM - Configuring for Price-Performance