Version 7.0 of IIS is included with Windows Server® 2008 and Windows. Windows Server 2008 R2 and Windows Server. Tools->Server Manager. You can install both IIS 6 & 7 in windows server 2008. Windows Server 2008 with IIS 6.0. You can have the management tools for IIS 6 but that is.
Sizing a Microsoft Server Farm Sizing a Microsoft Server Farm Determining Total Required Microsoft Servers Planning a Microsoft SoftGrid server farm begins with determining how many servers are needed to successfully support the Microsoft SoftGrid user community. Since Microsoft SoftGrid uses streaming to deliver applications, the basis by which we determine load is in streams as opposed to users. The formula below calculates the number of servers needed based on stream count. Total Required Servers = (Concurrent Stream Requirement / Individual Server Capacity) + Redundancy Factor Determining Concurrent Stream Requirement There is a slightly different model for calculating Concurrent Stream Requirement based on whether the SoftGrid Terminals Services Client or the SoftGrid Windows Desktop Client is being used due to the multi-user accessibility of the SoftGrid for Terminal Services client. For SoftGrid for Windows Desktop, the formula is Concurrent Stream Requirement = Concurrent Users * Concurrent Application Request For SoftGrid for Terminal Services, the formula is Concurrent Stream Requirement = Total Terminal Servers * Concurrent Application Request Concurrent Users This represents the maximum number of users who will be using the system at the same point in time. Concurrent Application Request This variable represents the number of applications assigned that are most likely to be launched back to back at any one time.
However, there is a difference in the variables when calculating load based on a SoftGrid for Windows Desktop Client vs. A SoftGrid for Terminal Services Client. In a SoftGrid for Windows Desktop model, each desktop is considered a single unit and the Concurrent Application Request is simply the typical number of applications a single user will launch back to back on their own PC.
For example, a user is deployed Outlook, Word, Excel, and a Line of Business application. Upon login they typically launch Outlook and the LOB application totaling 2 applications. The Concurrent Application Request is therefore 2. In a SoftGrid for Terminal Services model, each session will be launching applications from the same server. For example, a user is deployed Outlook, Word, Excel, and a Line of Business application. Upon login they typically launch Outlook and the LOB application #1 totaling 2 applications. However, in most Terminal Server cases, applications launched across the sessions are common to the workgroup and therefore the initial launch threshold is only streamed for the first user.
The subsequent users access the application from the SoftGrid cache without additional server streaming. Therefore, the true Concurrent Application Request would reflect the number of unique applications across all back to back launches per Terminal Server.
Based on our example, a typical user set might use the same e-mail program such as Outlook and a line of business application. The users would be launching the same two apps back to back, requiring only 2 streams per Terminal server. Revue Noir Anthology Archive Rapidshare.
The Concurrent Application Request for that session is therefore 2, however, there are multiple sessions across that terminal server. If half of those users launch LOB application #2 instead of LOB application #1, then there are now three unique applications being launched. Implementation Examples The following implementation examples depict three separate scenarios. First is a Microsoft SoftGrid implementation using only Direct Streaming, the second uses Microsoft SoftGrid as a management foundation for a Terminal Services implementation, and the third shows a “dual-mode” scenario with applications being streamed both directly to the end-user and to Terminal Services. Microsoft SoftGrid Direct Stream Implementation This system would be used to deploy applications to LAN and WAN users directly.