Architectures (IIS 6.0)

Applies To: Windows 7, Windows Server 2008, Windows Server 2008 R2, Windows Vista

Whether your server needs are simple with light loads or complex with heavy loads, a good understanding of the capabilities and techniques available to you in IIS 6.0 and Windows Server 2003 to access remote content using a UNC pathname will help you in implementing your design.

Distributed Scenario

As an example of a situation where IIS enables remote storage and distribution, consider an educational institution that has 100 educators producing schedules and papers, and posting grades for their classes (see Figure 1). IIS can be used to map to individual shared folders on each educator’s system, allowing them to quickly and easily publish important content to their students. In this case, IIS acts as a concentrator of distributed information using remote storage.

Figure 1: IIS 6.0 acts as a concentrator of distributed information using remote storage

This kind of a system could be made more reliable and manageable by using distributed file system (DFS) as a means to the underlying directory structure. For instance, if Anthro101 was mapped to a UNC pathname that was part of a DFS directory, when the instructor changes next semester, you need only change the association for \\Instructor\Anthro101 to the \\NewInstructor\Anthro101 in DFS. There is no need to change the IIS configuration.

ISP Scenario

IIS offers remote content benefits to Internet Service Providers (ISPs). For example, an ISP might have thousands of Web sites on a single server in a shared hosting environment. The content is spread out over many sites but traffic is low per site. By placing the content off the IIS server, the ISP can store content for multiple servers on a single RAID (Redundant Array of Independent Disks) device (see Figure 2). The ISP’s investment in storage is optimized by being concentrated instead of distributed to many Web servers. This reduces exposure to hardware failures, and simplifies recovery strategies.

Figure 2: IIS distributes information from a remote storage facility

High Volume Scenario

High-volume Web sites often use Web farms consisting of multiple, identically-configured IIS servers receiving traffic from a load balancer. In this scenario, it is common for the servers in the Web farm to access a cluster (or clusters) of Microsoft SQL™ servers, often using Microsoft Cluster Server (MSCS), as a common data store. Note that this cluster scenario does not use Network Load Balancing (NLB) between file servers because Server Message Block (SMB) communications used in Microsoft networks are session-dependent. This high-availability scenario provides the applications running on IIS 6.0 with a common database that has failover capability in the event of a problem with one of the servers. Similarly, the content and applications delivered by IIS 6.0 can be hosted on a clustered file server or NAS device, so that a failure of one of the remote stores is transparent to users (see Figure 3).

Figure 3: Multiple IIS servers provide their applications and content with a common database in a high volume or Web farm scenario

For more information about designing high availability server systems see: High-Availability System Architecture.