Flying High with Highly Available SQL Server Deployments
summary
Defines high availability for SQL Server as recoverable failover and shows a working product that pairs Windows Failover Clustering with DataKeeper to mirror local disks via block-level replication, avoiding SANs and enabling standard‑edition SQL instead of Enterprise. Demonstrates cloud support (EIPs, Route 53), Windows and Linux options, explains expected downtime for various “nines” (99% = ~3.65 days, 99.9% = ~8.76 hours, HA ≈ 52.5 minutes) and claims about 70% saving versus Enterprise licensing.
AI-written summary · about
chapters (10)
- 0:00High availability intro
- 1:12Defining HA importance
- 4:04HA in aviation parallels
- 5:58Application-level high availability options
- 6:40Data-level HA solutions
- 8:03Understanding 'nines'
- 10:49Why HA matters for SQL
- 12:56HA options for SQL Server
- 14:40SOS DataKeeper intro
- 17:28Q&A
related talks
20mFlying High with Highly Available SQL Server Deployments- 20mFlying High with Highly Available SQL Server Deployments
20mChallenge to Triumph: Exploring Real-World High Availability in SQL Server Environments- 20mChallenge to Triumph: Exploring Real-World High Availability in SQL Server Environments
52mMicrosoft SQL Server HA/DR on AWS
42mHigh Availability with SQL Server on Linux
49mEverything you always wanted to know about HA/DR for SQL Server on Azure VMs but were afraid to ask
52mBusiness Continuity for Hybrid and Cloud Native SQL Server Deployments