分布式应用程序

 

实现 .NET 远程处理包括以下阶段：

• 确定宿主应用程序或者环境。

• 创建服务器组件。

• 创建客户端。

• 编辑配置文件。

 

确定宿主应用程序或者环境

.NET 远程处理服务器组件需要在一个应用程序域中宿主，以监听该对象的传入请求。 宿主一个远程处理服务器组件您有四种选择：

• IIS 上的 ASP.NET

• 组件服务 (COM+)

• 一个 Windows 系统服务

• 一个 Windows 应用程序（基于控制台或 Windows 窗体）

您可以将 .NET 远程处理对象作为运行在 IIS 上的 ASP.NET 组件一样宿主。 这带来了许多好处，包括对安全性和易伸缩性的内置支持。 有了这些配置，您既可以用二进制也可以用 SOAP 格式化程序，但您只能用 HTTP 通道，它比 TCP 慢。

您可以将组件服务中的远程处理对象作为 Windows XP Professional SP1 或 winodws Server 2003 上的 COM+ 服务组件一样宿主。此配置提供了组件服务的企业功能，如集成的安全性、事务处理、池化和激活。

您可以在任何托管 Windows 服务中宿主远程处理对象。 在一个 Windows 服务中宿主一个远程处理对象为使用所选择的任何通道或者格式化程序配置提供了灵活性，包括 TCP 上的最高效率的二进制执行组合。 然而，.NET 远程处理并没有一个内置的安全模型。 因此，如果您要在一个 Windows 服务中宿主一个远程对象，则需要建立自己的身份验证和授权机制。

最后，您可以在一个 Windows 应用程序中宿主一个远程处理服务器组件。 然而，应用程序运行在登录用户的上下文中，因此需要用户登录后才能执行。 它们还运行在该用户的安全上下文中，这可能不适合您的环境。 在一个 Windows 服务中宿主一个远程服务器组件会更好。

注 当在一个 Windows 窗体应用程序中宿主一个组件时，一个 Windows 服务或者一些其他应用程序类型时，应由程序员指定端口。 对非 ASP.NET 主机来说，远程处理可以监听任何没有使用的端口。 

 

Partitioning applications into presentation, application logic, and data sections results in a simplified programming model that is the standard way to build applications that take advantage of intranet and Internet communications. By adding more hardware where needed, organizations can also expand n-tier applications to handle more users and information.

The n-tier model has proven especially important to developers wanting to integrate Web technologies into business processes. Because the model separates presentation, logic, and database functions, it is much easier to insert a Web server for backend processes and a client browser for presentation processes.

Taking Advantage of Hardware

The move to an n-tier development model is the result of a significant evolution of the personal computer application development model during the past decade. In the late 1980s, applications were generally written to run completely on a single PC. In the early 1990s, the two-tier client-server model emerged. This allowed developers to offload some of the more intensive data-processing work from client machines and move it onto more capable back-end servers. In this two-tier model, presentation software (the user interface) remained on the personal computer, while the bulk of the processing work was moved to the server. The three-tier model adds a further element of separation between the data and the presentation layers, by allowing processing logic to be handled on a separate server from the database and presentation functions.

Distributed applications are written to take advantage of the processing power available on servers—such as those used for database and line-of-business applications—and the client-side capabilities of personal computers and other devices. By dividing the processes used in an application among different computers, organizations can build flexible and scalable software using commodity hardware. 

see also :http://technet.microsoft.com/en-us/library/bb727077.aspx 
 

 

 

 

 Table 1 Cluster Nodes Supported by Operating System and Technology.

 

Operating System	Edition	Network Load Balancing	Component Load Balancing	Server cluster
Windows 2000
	Advanced Server	32	8	2
	Datacenter Server	32	8	4
Windows Server 2003
	Enterprise Edition	32	8	8
	Datacenter Edition	32	8	8

Scaling by Adding CPUs and RAM

When looking to scale up by adding CPUs and RAM, the edition of the server operating system used is extremely important.

In terms of both processor and memory capacity, Datacenter Server is much more expandable.

• Windows 2000 Advanced Server supports up to eight processors and eight gigabytes (GB) of RAM.
  
  
• Windows 2000 Datacenter Server supports up to 32 processors and 64 GB of RAM.
  
  
• Windows Server 2003 Enterprise Edition supports up to eight processors and 32 GB of RAM.
  
  
• Windows Server 2003 Datacenter Edition supports up to 32 processors and 64 GB of RAM.

• IT Staff Considerations

  As IT staff considers the impact of operating modes in their cluster architecture, they need to look carefully at the business requirements and the expected server loads.

  With NLB and CLB, all servers are active and the architecture is scaled out by adding additional servers, which typically are configured identically to the existing NLB or CLB nodes.

  With Server cluster, nodes can be either active or passive, and the configuration of nodes depends on the operating mode (active or passive), as well as how failover is configured. A server that is designated to handle failover must be sized to handle the workload of the failed and the current workload (if any). Additionally, both average and peak workloads must be considered. Severs need additional capacity to handle peak loads.

  Server cluster Nodes

  Server cluster nodes can be either active or passive.

  • Active Node. When a node is active, it is actively handling requests.
    
    
  • Passive Node. When a node is passive, it is idle, on standby waiting for another node to fail.
    
    

  Multi-node clusters can be configured using different combinations of active and passive nodes.

   
  

   

   
  

   see also:http://technet.microsoft.com/en-us/library/cc739634(v=ws.10)