Trends in Data Management: 'High Availability'

In today’s highly competitive business environment, for most companies it is imperative that their business data is always accessible from their data servers in a seamlessly fault tolerant manner. DBMSs, which constitute the core of most business information systems, are attempting to implement this requirement through a feature often referred to in the data management literature as ‘high availability’. Against the backdrop of a leading DBMS, Microsoft SQL Server 2005, we define the concept of high availability and examine the various approaches to high availability, namely, mirroring, clustering, replication, and log shipping, which should be of interest to database administrators and the students of database management system technology.