Windows azure Overview

ArticleDirectory

• Overview
• Computing
• Storage

Overview

Cloud computing has arrived. Run applications on machines in data centers that can access the InternetProgramAnd data storage, has many advantages. However, no matter where applications run, they are built on a certain type of platform. For internally deployed applications, such as applications running in an organization's data center, this platform usually includes an operating system, a data storage method, and other content. Applications running on the cloud need a similar Foundation.

Windows azure aims to provide it. As part of a larger windows azure platform, Windows Azure is the foundation for running applications and storing data in the cloud.

Without providing Microsoft customers with software that can be installed and run on their own computers, Windows Azure is now a service: the customer uses it to run applications and store data on all Microsoft machines that can access the Internet. These applications can provide services to enterprises, users, or both. Below are some examples of application types that can be built on Windows azure:

□An independent software supplier (isV) can create an Application Suitable for enterprise users, a method commonly referred to as software as a service (SAAS. Windows Azure is designed to support Microsoft's own SaaS applications to a certain extent, so isV can also use it as the basis for various enterprise-oriented cloud software.

□Isv can create a SaaS application for users rather than enterprises. Because Windows Azure is designed to support software with extremely high scalability, companies planning to enter the large consumer market can also choose it as a platform for new applications.

□Enterprises can use Windows azure to build and run applications for their employees. Although this situation may not require a large scale of user-oriented applications, the reliability and manageability provided by Windows azure still makes it an attractive choice.

To support cloud applications and data, Windows azure has five components ,:

These components are:

□Computing: run applications in the cloud. These applications largely see a Windows server environment,
However, the Windows azure programming model is different from the deployed Windows Server model.

□Storage: stores binary programs and structured data in the cloud.

□Structure Controller: deploy, manage, and monitor applications. The structure controller also processes system software updates throughout the platform.

□Content delivery network (CDN): accelerates global access to binary data in Windows azure storage
The ball maintains the cached copy of the data.

□Connection: allows you to create an IP-level connection between an on-premises computer and a Windows azure application.

 

Computing

Windows azure computing can run many different types of applications. However, no matter what operations the application performs, it must be implemented as one or more roles. Then windows azure usually runs multiple instances of each role and distributes requests to them using the built-in load balancing.

In the current Windows azure version, developers can select from three types of roles:

□Web role, mainly used to simplify the creation of web-based applications. Internet Information Service (IIS) is pre-configured in each web role instance. Therefore, it is very easy to use ASP. NET, Windows Communication Foundation (WCF) or other Web technologies to create applications. Developers can also use the Local MachineCodeCreate an application -- you do not need to use. NET Framework. This means they can also install and run non-Microsoft technologies, including PHP and Java.

□Job role, designed to run different Windows-based code. The biggest difference between Web roles and worker roles is that IIS is not configured in the worker color, so the code they run is not hosted by IIS. For example, a worker role can simulate, perform video processing, or perform almost any other operation. Applications often interact with users through web roles, and then forward tasks to the worker roles for processing. Again, developers are free to use. NET Framework or other software running on Windows, including non-Microsoft technologies.

□Vm role, each of which runs a user-provided Windows Server 2008 R2 image. In addition, Vm roles are sometimes useful for migrating internally deployed Windows server applications to Windows azure.

To submit an application to Windows Azure, developers can use the Windows azure portal. In addition to applications, she also submits configuration information to tell the Platform How many instances of each role to run. The Windows azure schema Controller then creates a Virtual Machine (VM) for each instance and runs code for the appropriate role in the VM.

 

Storage

Applications process data in many different ways. Accordingly, the Windows azure storage service provides multiple options. These options are provided.

The simplest way to store data in Windows Azure is to use blob. Blob contains binary data, and it can be seen that the hierarchy is simple: each container can contain one or more blobs. Blob can be large (1 Tb) or contain associated metadata, such as information about the location where the JPEG photo is taken or the artist of an MP3 file. Blob also provides underlying storage for Windows azure drives, which allows Windows azure role instances to interact with persistent storage, just as it is a local NTFS file system. Blob is only applicable to some situations. In other cases, they are too unstructured. To allow applications to process data in a more fine-grained manner, Windows azure storage provides tables. Do not be misled by names: these are not Relational Tables. The data contained in each table is actually stored in a group of entities, and the entity contains attributes. Without SQL, applications can use the conventions defined by odata to query table data. This method supports more effective storage scaling than standard relational databases (by spreading data across multiple machines ). In fact, a Windows azure table can contain billions of entities and store several terabytes of data. Blob and table are mainly used to store and access data. The third option in Windows azure storage is queue, which has different purposes. One of the main functions of queue is to provide a way for Web role instances to communicate asynchronously with the worker role instances.

For example, you can submit a request to execute a computing-intensive task through a Web interface implemented by the Windows azure web role. The Web role instance that receives this request can write a message to the queue describing the task to be completed. Wait for the worker role instance of this queue to read the message and execute the task it specified. Any result can be returned in another queue or processed in some other way.

No matter how data is stored (blob, table, or Queue), all information stored in Windows azure storage is replicated three times. This type of repetition achieves fault tolerance because losing a copy will not cause a fatal impact. However, the system provides strong consistency, so the application that immediately reads the data it just writes can ensure that it can get the content it writes.
Windows azure also saves one backup copy of all data in another center in the same location. If the data center that saves the primary copy is unavailable or damaged, you can still access the backup.

Windows azure storage is accessible to Windows azure applications, internally deployed applications, or applications running on one managed platform or another cloud platform. In all these cases, all 3 windows azure storage styles use rest conventions to identify and expose data, as shown in figure 4. Blob, table, and queue are named by Uri and accessed through standard HTTP operations .. The. NET client allows the library provided by Windows azure to perform this operation, but the application can also execute the original HTTP call without doing so. It is undoubtedly useful to create a Windows azure application sequence that uses blob, table, and queue. Applications dependent on relational storage can use SQL Azure, another component of the Windows azure platform. Applications running on Windows azure (or elsewhere) can use this technology to access relational storage in the cloud for SQL-based familiarity.

From author David Chappell, Windows azure Overview

Author: flute)

Source: http://www.cnblogs.com/flute-di/

the detailed link of this article must be provided during reprinting; otherwise, the author will be held legally liable.