Component blocks Windows Azure uses to build powerful cloud applications

Following the publication of a design-strong cloud Application blog, the http://www.aliyun.com/zixun/aggregation/13357.html ">azure CAT Team Program provides code projects for publishing on the MSDN code base Detailed information and technical Instructions for components in the cloud services infrastructure in Windows Azure. We will launch a series of blogs and technical articles to illustrate the use of these basic building blocks, which we call components. In the next few months, we will release a series of blogs every other week in Thursday, providing detailed technical instructions on the various components of the cloud service base.

Over the years, we have been working to collaborate with Windows Azure customers both inside and outside of Microsoft, and to delve into the needs of Windows Azure services on several occasions. We have also learned that some seemingly fundamental questions about how to implement cloud services are becoming increasingly complex. For example, we need a data access layer, not just a section of partition code. In order to realize the flexible data access layer, we need to develop the retry logic. We also need a reliable guide on how to record a large number of errors. Let alone build an OPS store that can query reports and generate alerts. You can see how the discussion progresses with each component because they depend on each other. These discussions and implementations generate code projects The cloud services infrastructure in Windows Azure links Some of the basic components to a functioning cloud application.

For the CAT team, this code project is a huge challenge because we are committed to implementing complex, database based services on Windows Azure for the largest number of customers. It is based on our efforts to solve the specific problems that Windows Azure customers are experiencing. When we take into account the many requirements of a large cloud service (including flexible scaling, partitioned workloads, availability, business continuity, a large number of distributed users, and high capacity low latency requests), solving these problems is not as simple as relying on basic examples, but often requires best practices. You can view the infrastructure of the cloud services underlying code project below.

Our series of technical articles will provide a detailed description of the components in the code project, including:

1. Remote control – It is the basis for measuring and documenting application services through an asynchronous mechanism implemented on a large scale in a data pipeline. Effective use of remote control data is the key to troubleshooting services and determining service performance. The code project uses the background worker role implementation Scheduler to periodically collect remote control data about applications, performance counters, IIS logs, event logs, and the DMV for partitioned SQL databases. A custom OPS storage database that writes data to a Windows Azure SQL database. The data collected by the scheduler can be viewed through reports in SQL Reporting.

2. Data Access Layer – the ability to efficiently and reliably access multiple databases in a Windows Azure SQL database. The code project has a data access wrapper for a single database and partition solution, and displays techniques such as cross section parallel fan out queries.

3. Caching – By combining Azure caching with the data access layer, you can more efficiently store and retrieve user data from a private cache.

4. Configuration – The configuration file is the key to implementing seamless management of the application, regardless of whether the configuration parameters are located in Web.config or service.config– this should be transparent to the application.

5. Application Request Routing-leverages the ARR (Application request routing) technology in IIS, implements Cookie Routing and associates users to multiple managed services and partitioned databases to extend the application service level to a partitioned database.

6. Deployment – A way to deploy custom configurations to multiple managed services and variable number of instances and to configure the number of partitions by using the Windows Azure Cmdlet in PowerShell.