《Chapter13ExploitingSQLAzureDatabasesRelational》.docVIP

Chapter 13: Exploiting SQL Azure Databases Relational Features Cloud-based relational database management systems (RDBMSs) are relatively uncommon because they are difficult to implement with the enterprise-grade scale-out and availability capabilities IT departments have come to expect from traditional schemaless Entity-Attribute-Value (EAV) data stores, such as the Google App Engine’s datastore and Amazon Web Services’ SimpleDB. Google offers an SQL-like query dialect called GQL and Amazon provides a set of SimpleDB API calls to emulate simple SQL-like SELECT statements. Neither of these query “languages” approaches the capability of ANSI SQL to deliver complex result sets tailored to a particular structure and having a precise, predictable composition. These tables also lack the notion of relational INNER or OUTER JOINs, but the App Engine’s datastore can emulate many:one associations with db.Key and one:many associations with list(db.Key) data types. type=note The “Comparing Google App Engine, Amazon SimpleDB and Microsoft SQL Server Data Services” post of May 6, 2008 to the OakLeaf Systems blog (http://bit.ly/LtvxN, /2008/04/comparing-google-app-engine-amazon.html) describes differences in the capabilities of these three EAV datastores at the time. SQL Server Data Services (renamed SQL Data Services when this book was written) will be retired when Microsoft releases SQL Azure as a commercial product in late 2009. Before you can query relational or EAV cloud-based tables, you must specify a relational schema or define entities and their properties, as well as populate tables or entity collections with data. Most early cloud databases or datastores will be based on schemas of on-premises databases and contain at least a nominal amount of historical data. Undoubtedly, getting data-intensive applications up and running in the cloud will be faster and simpler with an RDBMS rather than EAV tables as the datastore. This chapter covers migrating schemas and data from