Databases are at the heart of every organization. They support line-of-business applications, internet applications, human resource systems, financial applications, and more. Application downtime and database errors mean lost customers in both the short- and long-term. According to a 2007 University of Warwick study, hourly downtime costs were pegged at $682,000 per hour in the retail sector and $192,000 in the financial sector. Those costs are probably conservative and could be higher or lower, depending on the size and nature of your business. What are the causes? Causes include such factors as natural disasters, hardware failure, upgrades, human error, and others. Achieving high availability is a critical component of our job as database administrators (DBAs). Your final high availability (HA solution) will depend on what you are trying to protect against, what is your acceptable downtime tolerance, and what will operate along side your Disaster Recovery strategies.
SQL 2005 introduced high availability and scalability enhancements, such as database mirroring and improved failover clustering from earlier versions. SQL Server 2008 has built on the earlier SQL 2005 architecture and added some interesting new features. Microsoft is targeting close to zero downtime with SQL 2008 with availability and recovery features that only used to be available via third-party solutions. These features include prioritization management, hot-swap replication, and the ability to recover off of a mirrored server. This white paper discusses the evolution of High-availability from SQL 2005 to SQL 2008, and the tools available to minimize downtime in the Enterprise.
SQL Server Research Library
