SQL Server on Azure VM - Configuring for Price-Performance
summary
Covers concrete configuration choices: choose ebds V5 (double‑sprinkles) for mission‑critical workloads and Azure Boost, use premium ssv2 for most OLTP/IO‑intensive disks, put tempdb on the ephemeral NVMe, set host caching to read‑only for data and none for transaction logs, and watch IO caps — examples include a 112‑core preview target of 400,000 IOPS and 10 GB/s throughput and a common VM limit of 12,800 IOPS; use the SKUR recommendation tool via Azure Data Studio (aka.ms) and measure iops/throughput/latency first.
AI-written summary · about
chapters (11)
- 0:00SQL Server on Azure VM
- 1:53Azure VM sizing and scaling
- 7:58Azure VM families for SQL
- 11:06Storage basics for SQL
- 22:07Workload analysis tool
- 22:49EBDs V5 and premium SSD V2
- 25:01Storage for SQL workloads
- 33:42NVMe vs SCSI performance
- 38:42IO throttling analysis
- 43:44Host caching best practices
- 48:01Best practices and closing
related talks
42mSQL Server on Azure VM - Configuring for price-performance
54mAzure Best Practices for Lift and Shift
50mSQL Server to Azure SQL: Performance and Availability
59mWhat's new on Azure IaaS for SQL Server VMs
53mOptimizing your Azure SQL Database workloads
01h 01mSQL Server in the Cloud - The available options
01h 05mHow to Get the Best Out of Your SQL Server on Azure
18mHow to reduce your SQL Database costs on Azure
