Book Image

Microsoft Windows Azure Development Cookbook

By : Neil Mackenzie
Book Image

Microsoft Windows Azure Development Cookbook

By: Neil Mackenzie

Overview of this book

The Windows Azure platform is Microsoft's Platform-as-a-Service environment for hosting services and data in the cloud. It provides developers with on-demand computing, storage, and service connectivity capabilities that facilitate the hosting of highly scalable services in Windows Azure datacenters across the globe. This practical cookbook will show you advanced development techniques for building highly scalable cloud-based services using the Windows Azure platform. It contains over 80 practical, task-based, and immediately usable recipes covering a wide range of advanced development techniques for building highly scalable services to solve particular problems/scenarios when developing these services on the Windows Azure platform. Packed with reusable, real-world recipes, the book starts by explaining the various access control mechanisms used in the Windows Azure platform. Next you will see the advanced features of Windows Azure Blob storage, Windows Azure Table storage, and Windows Azure Queues. The book then dives deep into topics such as developing Windows Azure hosted services, using Windows Azure Diagnostics, managing hosted services with the Service Management API, using SQL Azure and the Windows Azure AppFabric Service Bus. You will see how to use several of the latest features such as VM roles, Windows Azure Connect, startup tasks, and the Windows Azure AppFabric Caching Service.
Table of Contents (16 chapters)
Microsoft Windows Azure Development Cookbook
Credits
About the Author
About the Reviewers
www.PacktPub.com
Preface
Index

Performing queries in parallel


The performance of a query against a table in the Windows Azure Table Service is critically dependent on whether or not the PartitionKey and RowKey are provided. The only index on a table is the combination of PartitionKey and RowKey. The most performant query, and the only one guaranteed not to result in a continuation token, is one specifying both PartitionKey and RowKey. A query specifying the PartitionKey but not the RowKey results in a scan of all the entities with that PartitionKey. A query specifying the RowKey but not the PartitionKey is processed sequentially by each partition server. A query specifying neither PartitionKey nor RowKey results in a scan of the entire table regardless of any filters in the query.

The Table service provides lower scalability targets for simultaneous queries within a single partition than it does for simultaneous queries against different partitions. Consequently, it can be more efficient to break a multipartition query...