Consistency of Reads in Microsoft SQL Server
Play Consistency of Reads in Microsoft SQL Server
When you design an application that reads data from SQL Server you need to balance performance and consistency. Your index design and choice of isolation level will have an integral impact on performance and consistency of your queries. There are some aspects of indexing and choice of isolation level that are of common knowledge; however, there are some recent discoveries (in the last year or two) that could make you rethink your choices. For example, would you expect a SELECT query to return the same row multiple times? Would you expect a SELECT query to skip a row altogether? This session explains the circumstances where such consistency problems can happen, and how to prevent them.