Considering the Windows Server 2008 Editions

Considering the Windows Server 2008 Editions

Windows Server 2008 comes in a number of versions and a number of editions. A version affects functionality in some way. For example, if your server requires a 32-bit operating system, you use the 32-bit version of the product. Likewise, anyone with an Itanium processor will require the Itanium version of the product. Windows Server 2008 breaks with tradition in that it also offers a Server Core version.

This new lightweight version is Windows without the windows; you can read more about it in the “Understanding Windows Server 2008 Server Core” section of this chapter. The Server Core version comes in Standard, Enterprise, and Datacenter editions. You can’t get Server Core in the Web or Storage Server editions because these editions require the use of a graphical interface and Server Core doesn’t provide the required support.

The Itanium Processor version is specifically designed to run on the Itanium processor. Except for not supporting a 32-bit version (because this processor doesn’t come in a 32-bit format), the features of the Itanium Processor version are much like those found in the Enterprise edition.

An edition defines the feature set of a particular version of the operating system. The Standard edition offers a different feature set from the Enterprise edition. You can get both editions in 32-bit versions. Consequently, you need to know which version and which edition you want before you purchase Windows Server 2008.

Although figuring out which version you want is relatively easy because you already know about your hardware, figuring out the edition is a lot harder because you have to consider features that you may not know about. The following sections describe each of the editions that Microsoft is offering for Windows Server 2008.

Memory considerations

The edition you choose determines how much memory the server will support. Sometimes, the memory limits aren’t a problem. If you have a small company with basic needs, you’ll never exceed the limits of even the 32-bit version of the Standard edition. However, if you’re working in a large enterprise and your server is part of a server farm, you’ll probably need something more robust than the Standard edition. Table 1-1 provides you with the memory limitations for various editions of Windows Server 2008.


Table 1-1 Memory Limits for the Windows Server 2008 Editions

Other hardware considerations

You’re going to need a relatively large hard drive for Windows Server 2008. Even though Microsoft says that you can get by with a 10 GB hard drive, that’s hardly realistic. Even the 40 GB hard drive that Microsoft Recommends is quite small for a Windows Server 2008 setup. A real setup is going to have as much hard drive space as you can afford. Because hard drives are inexpensive today, you should consider getting a 500 GB or larger hard drive and make sure to include a Redundant Array of Inexpensive Disks 1+0 or 0+1 setup in your plans, for reliability reasons.

You can read more about the various RAID levels at http://en.wikipedia.org/wiki/Standard_RAID_levels,In addition to having hard drive support, you must have a Video Graphics Array (VGA) display adapter (or better) capable of at least 800 x 600 resolution. It’s important to remember that this is a server, not a workstation. In many cases,you won’t need a super-high-resolution display adapter and monitor for a server. In fact, in some cases, it’s detrimental to provide too much screen real estate because moving all those bits around will simply consume processing cycles you could use for some other task.

You normally need a mouse too when working with Windows Server 2008. It’s possible to get by without a mouse by using hardware shortcuts, but you may find that some tasks become difficult when you go this route. As with the display adapter, you don’t need a fancy mouse to work with the server. 24 Considering the Windows Server 2008 Editions

Standard

The Standard edition provides functionality that the average work center, small business, or even medium-size business requires.. The main limitations for the Standard edition are the number of processors (up to four processors in a single server) and the amount of memory you can use. (See the “Memory considerations” section of this Post for details.)

Enterprise

The Enterprise edition provides functionality that a medium-size business will normally require for a centralized server. You also find this edition in larger businesses as part of a large workgroup. In addition to all the functionality that the Standard edition provides, the Enterprise edition also provides greater memory support and the ability to use more processors (up to eight processors in a single server). Additional features include clustering and hot-add memory support.

Datacenter

The Datacenter edition provides support for large businesses as part of the centralized server setup. It provides the maximum processor (up to 64 processors in a single server) and memory support allowed by Windows Server 2008. You receive, in addition to the features found in the Enterprise edition, improved clustering support and the ability to dynamically partition hardware. Additional features include hot-replace memory and hot add/ replace processor support.

Web

The Web edition is a special Windows Server 2008 setup for the sole purpose of supporting Web sites. Consequently, you won’t find many of the features found in other editions in the Web edition. This edition focuses on .NET Framework, IIS 7, and ASP.NET support. The reason you want to get this edition for Web sites is that it uses resources more efficiently and helps you support Web sites with fewer resources. In addition, having fewer features translates into a more reliable and secure server because there are fewer ways in which the server can fail. The memory and processor limits for the Web edition are the same as for the Standard edition. You can’t use the Web edition to create virtualized images.

New Features in Windows Server 2008

Network access protection (NAP) and enforcement

Every time other people access your network, they interact with it in ways that could cause contamination on their systems to also appear on your server. For example, if a laptop user gets a virus, your server will very likely get the virus too, unless you have a number of safeguards in place. Of course, most organizations today have all kinds of spam, virus, and other types of protection in place, even for internal contacts. The problem remains one of unhealthy clients — clients who lack the required updates or have some other fault that makes them a risky connection.


Network Access Protection (NAP).

NAP provides the means to check the health of any client connecting to your
network. When NAP detects an unhealthy client, it can act by limiting access until the client receives the necessary updates and configuration changes. In addition, you can specify a means of fixing the client to place it in a healthy state again. The bottom line is that your system gains another layer of protection from outside influences.

New printer and storage options

Microsoft has provided a number of new printer and storage options in Windows Server 2008. Most of these features fall into the required upgrade category. For example, Microsoft really needed to provide a means to access devices from the Web, so it provided a means to do that. Administrators have complained for years about the management tools in Windows, so Microsoft has also addressed that requirement. The following list provides some details on these new features:

✦ XML Paper Specification (XPS) Document Support:
XPS is a new open document format that Microsoft is promoting. It relies on XML to store document data so that you can theoretically retrieve the data even if the application used to create it no longer exists.

✦ New print paths:
A print path defines the software and actions that the operating system uses to process a document and send printer-ready data to the printer for output. In the past, the only print path relies on the Graphics Device Interface (GDI) originally found in Windows 3.x (albeit modified with each version of Windows). Microsoft now provides the XPS Driver (XPSDrv) software to process XPS documents more efficiently than ever. This new print path includes a host of features, such as direct support for transparencies, but you have to have the WPF installed to get it . You find a number of other useful additions to Windows

Server 2008 in support of the XPS specification.

✦ New printer driver model:
XPSDrv software represents a new way of outputting data. However, it also supports older output methodologies using the new print path. Even though XPSDrv software provides support for newer technologies, such as WPF, you can still use the driver without the .NET software. Of course, you won’t get any of the new features, but you will gain the performance benefits.

✦ Scalability improvements:
Normally, when a client makes a print request, the server performs all the required processing. Of course, placing the burden on the server reduces overall performance of the system for a very small gain in network performance. Windows Server 2008 now places the burden of printing on the client.

✦ Web Services on Devices (WSD):
This new feature relies on a common framework to describe and manage network devices. The concept is the same as a Web service, where you can query the device for specific information by sending it an XML request. Because this technology hides the details of working with the device, it greatly simplifies device interaction and reduces the potential for error.

✦ Improved print server management tools:
Anyone who has used Windows 2003 R2 has to appreciate the Print Management Console (PMC). Microsoft has improved this tool in Windows Server 2008. For example, you can now use PMC to migrate a printer from Windows 2000 or Windows 2003 to Windows Server 2008. The suite of tools also includes an improved Printer Installation Wizard that reduces the number of steps that an administrator must perform to detect and install a printer. As you can see from the list, Microsoft has provided some necessary and some nice-to-have printer and storage features in Windows Server 2008. It’s important to note that you still have to work pretty hard to obtain drivers for some hardware, but at least the support will be there after you find the drivers.

Read-only domain controller (RODC)

Even though you consider your domain controller safe from external influences, someone could potentially break into it and make changes to your setup. In addition, errant applications can cause havoc to your system. These are just two of the reasons for having a read-only domain controller (RODC).

Using an RODC makes Active Directory a read-only environment, which means that any changes that anyone makes are automatically overwritten with the original data. No changes occur to the server unless you make them. The main reason to use an RODC is that you can’t guarantee the physical security of the domain controller or have other concerns about the maintenance of server data.

For example, the RODC could appear in a branch office where a less-skilled administrator performs application installation or other tasks. In most cases, this less-skilled administrator won’t have access to any other server and won’t appear as part of the Domain Administrators group. Don’t worry that an RODC automatically becomes outdated either.

Unlike with a writeable domain controller, you can’t make changes directly to Active Directory, but the server can make requests for data from a writeable domain controller. For example, when a user tries to authenticate against thedomain controller the first time, the RODC sends the authentication request to a writeable domain controller.

If the authentication is successful, the RODC can request a copy of the credentials from the writeable domain controller and store them locally. Using an RODC improves security by making it impossible to write data to the local copy of Active Directory. However, it also provides benefits to the branch office. A special feature allows non administrators to gain a temporary increase in privileges to install applications and perform some limited management tasks so that an administrator doesn’t have to continually run to the branch office. In addition, because the RODC provides a local cache of Active Directory data, users will notice a performance increase because the network doesn’t continually experience the latency issues found in a standard

wide area network (WAN).

Windows Deployment Services Microsoft wants to make it easier for you to get its product on your machine, and Windows Deployment Services (WDS) is another in a long line of features designed to make that happen. The theory is that you use this new service to deploy Windows over a network on systems that have no operating systems installed. This feature originally appeared in Windows Server 2003 as a separate install, but you get it as part of Windows Server 2008. One especially nice addition is that you can perform the deployment without being physically present at the remote machine. You also find these features included with the new WDS in Windows Server 2008:

Enhanced Trivial File Transfer Protocol (TFTP) performance:

Windows Deployment Services relies on TFTP to download network boot programs and the Windows Pre-installation Environment (PE) images to the client.TFTP now includes a configurable windowing mechanism that reduces the number of packets that clients send, improving performance.

Diagnostics:
Windows Deployment Services now provides a considerable amount of information about the client install in the form of Crimson logs. You can use any Crimson-compatible product, such as Microsoft Office InfoPath, to process the logs and perform data mining on them.

Multicast deployment:
Using a multicast deployment technique lets the server install Windows on more than one client at a time using the same transmission thread. This approach improves overall performance and makes it possible to update the clients quickly. You have a choice between ScheduledCast (a task-based deployment where the client must join the session before the start of the download) or AutoCast (available at any time) deployment. Windows Server 2008 provides monitoring so that you can see how much of the multicast deployment the individual clients have received.

Internet Protocol version 6 (IPv6)

The world is running out of Internet Protocol (IP) addresses because every device seems to require one these days. When IP version 4 (IPv4) originally appeared on the scene with 4,294,967,296 possible address combinations, the standards groups thought that no one could ever use that many addresses. Unfortunately, the standards groups were wrong, and we now need IPv6, which provides a significantly larger address space of 3.4028236692093846346337460743177e+38 possible addresses. Of course, the standards groups also took this opportunity to improve performance, add support for mobile devices, and make IP more secure.

Just because Windows Server 2008 supports IPv6 doesn’t mean that everyone supports it. Before you can gain the benefits of IPv6 on your own network, you need to update older machines to use it. Microsoft has no plans to support IPv6 on Windows 2000 or Windows 98 machines, as outlined in the FAQ at http://www.microsoft.com/technet/network/ipv6/ipv6faq.mspx. The same FAQ tells you how to install IPv6 support for Windows XP and Windows 2003. Fortunately, both Vista and Windows Server 2008 include IPv6 as part of their default configurations. Adding IPv6 to your own network won’t do anything for external connections. To gain the full benefits of IPv6, you must also encourage your Internet service provider (ISP) to upgrade, which may be more difficult than you think. At some point, you can expect IPv6 to become a reality across the Internet, but the wait may be a long one.


.NET Framework 3.0

The .NET Framework is now part of Windows when you install it. The main reason for this change from previous versions is that Microsoft is using the .NET Framework more and more within Windows applications. For example, you’ll find that the Internet Information Services (IIS) Manager relies on the .NET Framework. In fact, you may very well find that many parts of Windows Server 2008 rely on the .NET Framework, which means that you have to get used to some new rules for setting security. The .NET Framework 3.0 isn’t much different from the .NET Framework 2.0 in many respects — at least, not in ways that you’d notice immediately. The main difference is that this new version of the .NET Framework provides access to some new features that Microsoft includes in both Vista and Windows Server 2008. These features appear as part of new foundations that Microsoft provides for making development easier, as shown in the following list

✦ Windows Presentation Foundation (WPF): Lets you create interesting graphics displays. In general, this feature applies only to client machines because it provides access to features such as the Aero Glass display in Vista. A server developer could possibly use this feature to provide graphs, charts, and other graphical elements in an application, but it’s unlikely.
✦ Windows Communication Foundation (WCF): Provides access to Web services. A Web service is a special kind of external connection to resources on another machine. For example, an application on your server could request a database update from another server through a Web service. In most cases, you won’t even know that the application is using a Web service to perform its work, because everything happens in the background.

✦ Windows Workflow Foundation (WF): Provides a means of creating an application that relies on a workflow to perform a long-running task. Workflow is a new marketing term from Microsoft that essentially means performing a task using a specific process. Using this feature, a developer can create an application where activities must proceed in a specific order based on various conditions, such as the successful completion of another
activity. Using workflows tends to reduce human error, which is the entire point of using this functionality.

✦ Windows CardSpace: Maintains user digital identity information in a secure environment that provides an ease of use equal to working with identity or credit cards. These virtual identity cards make it easier to gain access to resources online. In addition, because the user doesn’t actually have to remember much, the digital identity can rely on complex passwords and other means of identifying the user to the server. Generally speaking, all the existing .NET applications you have will run just fine under Windows Server 2008. About the only time you notice a change is if the application requires a new .NET Framework 3.0. In most cases, these changes appear only under the new operating systems.

Internet Information Server (IIS) 7

Internet Information Server (IIS) 7 is so completely different from what has gone in the past that you probably won’t recognize it. The interface is different, the configuration is different, and the internal workings are different. The reason for the massive change is that the old IIS just wasn’t keeping up with the latest Web technologies. In addition, at least some of the problems that administrators experience with IIS are due to the older design.

The biggest difference you’ll find in IIS 7 is that it places a new emphasis on administrators working with developers to create a Web solution. You can now configure ASP.NET applications in ways that you could only imagine in the past. Developers can also include new settings in applications to make the applications more responsive to enterprise needs.

Microsoft has also placed an emphasis on .NET development in IIS 7. Yes, you can still provide static content, and most scripted applications will work as well as they did before.The result is that your ASP.NET applications will perform better than ever.

IIS 7 includes a wealth of other changes that you’ll find make your job easier. For example, all the configuration files are now based on XML instead of relying on an arcane database that’s prone to corruption and other problems. You can configure IIS easily from the command prompt, from within the Internet Information Services (IIS) Manager, or by using a simple text editor to change the files directly.

Federated rights management

Active Directory Federation Services (AD FS)

Windows Server 2008 includes a number of new roles, one of which is Active Directory Federation Services (AD FS). This feature lets an administrator define an access identity across a network even if the access occurs over the Internet. In addition, this feature relies on the role-based security built into the .NET Framework.

Consequently, when someone logs in to the system, they have the rights defined by their role. A manager may have only manager rights when logging in from a local system — they may actually appear in the user role logging in from the Internet.

The AD FS role works across platforms, so it no longer matters if your network contains a mix of Windows, Linux, and Macintoshes. The administrator can also provide role-based authorization to Windows Share Point Services (WSS) and Rights Management Services (RMS) for a federated partner. A new Group Policy feature lets the administrator limit federation service deployment. It’s also possible to check on certificates by using the certificate-revocation checking settings.

Rights Management Services (AD RMS).

A second Windows Server 2008 role is Active Directory Rights Management Services (AD RMS). This service provides the means to attach usage rights to the data on your server. The rights to that data remain persistent no matter where someone moves it. You would use this feature to provide security for sensitive documents, such as financial reports.

The combination of ADFS and ADRMS form the Federated Rights Management feature. The use of these two server roles together provides a package of persistent, secure data management.

Improved failover clustering

Failover clustering is an important feature for a multi server network, where one server can take over for another when a failure occurs. Of course, you have to have the right hardware and software to create a clustered network, which means having knowledge of precisely what Windows is looking for in a cluster. Unfortunately, it was very difficult to make this determination in the past.

Microsoft has fixed this problem in Windows Server 2008 by providing validation tests you can use to ensure that your setup will work as a cluster. You now have access to node, network, and storage tests that determine whether a cluster will work and provide you with tips on resolving any potential issues.

After you determine that your hardware will work, you need to perform the setup. The new cluster software performs a configuration validation before it attempts to install the cluster. When the validation passes, the administrator can use a single-step setup process to install the cluster.

Daily maintenance is a requirement for any clustered setup. The new software provides easy methods for adding and removing clustered resources as needed. It’s also possible to perform management tasks from the command line (using standard cluster utilities) or with Windows Management Instrumentation (WMI). In fact, you can combine the two and use the command line WMIC tool to work with WMI at the command line.

The biggest improvement in failover cluster management is the way in which you can interact with hard drives. For example, you can now add a hard drive resource while the cluster is serving applications. The actual hard drive interactions are also improved. Microsoft has made changes that improve performance, such as not relying on SCSI hard drive resets. In addition, the software no longer leaves the hard drives in an unprotected state, which reduces the risk of corruption. You can even use a GUID Partition Table (GPT), as contrasted with the standard master boot record (MBT) partition, for increased hard drive space and reliability

An Overview of Windows Server 2008

Installation and Setup

Microsoft tries to improve each version of Windows Server. Most new versions offer improved reliability, performance, and security. Windows Server 2008 also includes a new version of Internet Information Server (IIS) that’s so different from what you used in the past that you might not even recognize it as the same product. In the new Windows Server 2008 Core Services, this is a Windows without windows. All you get is a command prompt with this version of Windows.

An Overview of Major New Features in Windows Server 2008

Bit Locker drive encryption

Bit Locker, a new feature in Vista, has also made its way into Windows Server 2008. This particular feature has the potential to improve system security immeasurably when it comes to tampering. Because Windows encrypts the entire hard drive, anyone attempting to read the hard drive outside the server won’t get very far. In addition, someone can’t even start the server without the required information because the hard drive encryption keeps the data completely locked (including the Windows boot code) until you provide the required code.

This feature makes a lot more sense on a laptop or other machine that leaves your premises regularly. Someone leaving a laptop sitting in a public location won’t then compromise all that precious data you worked so hard to accumu¬late. However, using BitLocker does mean that you don’t have to worry as much about someone gaining access to company data when you perform an upgrade of the hardware. Even if your assistant forgets to wipe the hard drive, no one can get to the data at the recycling center.

Enhanced Windows Firewall

Microsoft has been working hard to improve the security of Windows. The Windows Firewall in Windows Server 2008 is considerably easier to use than in previous versions of Windows, and it provides additional functionality. One method it’s employing is to make it harder for outsiders to gain entry to the server while making it easier for the network administrator to perform the required configuration. The initial display says it all by providing you with a quick indicator of firewall status , as shown in Bellow

The new firewall makes it easy to detect its current status. Configuring the firewall is considerably easier than in the past.

Understanding and Installing Cabling

Understanding Methods of Network Cabling

Network cabling is a method of connecting two or more computers so that they can communicate and share data, programs, and peripherals. Whether you’re cabling a small or a large network, you can use traditional cabling methods, which consist of various wiring and hardware techniques. With the continuing popularity of home networking, many manufacturers are producing networking kits that make cabling faster and easier. Additionally, techniques of using home phone lines and electrical housewiring (or power lines) are changing the scope of home networking altogether. Wireless connections are another popular option. Often various methods of networking cabling are combined.

The type of cabling you use in your network depends on a number of elements. You should consider network cost, the speed of the network connections, available resources, and so on. This Post and the next two explain the various methods you can use to connect your computers as a network.


Considering Networking Topologies and Technologies

You must lay certain foundations when building a network. In addition to choosing whether the network will be of the workgroup (or peer-to-peer) or client/server type, you should choose which topology and technology you will use for the network.

The topology of the network refers to how you arrange the cables, the networking hardware, and the computers. Technology refers to the type of wiring and hardware you use and to the general speed of the network. Network speed refers to how fast the data is transferred between two or more computers.

Basically, four topologies are possible:
✦ Bus
✦ Spanning tree (or star)
✦ Ring
✦ Mesh


Understanding Cabling
The cabling is a fundamental part of networking. Cabling provides the physical connection between computers; cabling is used for transmitting and receiving information over the network. You can connect your network with any of various types of cabling, or wires.

CAT 5 is the best cable for any network—business or home. CAT 5 works equally well with 10Base-T or with 100Base-T. You might want to start your network with CAT 5 cabling and 10Base-T hardware—10 Mbps network cards and hub—and then when you’re ready, upgrade to 100Base-T hardware. 100Base provides transfer rates of 100 Mbps.

Using twisted-pair cable connectors
Twisted-pair cabling uses different connectors than coaxial cabling uses. In twisted-pair, you use one connector to plug into the network card or hub. You can use a second connector to plug into a wall panel, called a jack, or patch panel to help keep the cables safe from kinking and to look nicer in your home

Teaching the use of the network

Teaching the use of the network

Before you can expect everyone to use the network without any problems, you’ll have to teach them the finer points of network usage. You may want to sit down with the entire family or go over it with each individual. It’s a good policy to show the adults in the house how to use the network first, and then let them help you teach the kids.

You don’t have to teach everyone the intricacies of the network; you just need to show them how to get what they need and go where they must go. The less you tell kids, as a matter of fact, the better off you’ll be in the beginning. As teenagers work with the network more and more, you can begin to show them the details of how it works.

Consider the following issues to teach the kids:

✦ Browsing the network
✦ Directory structures
✦ Opening folders on other computers
✦ Copying folders and files from others’ computers
✦ Creating folders on others’ computers
✦ Backing up to a specific drive or folder
✦ Finding the printer or other peripherals
✦ Choosing the default printer (if you have more than one printer)

Server types

Server types

If a network uses only one server, the server provides all services to the clients. Some services include file storage, application management, printer and other peripheral management, backup and restoration services, Internet access, and so on. If not many clients are attached to the network and the server is a powerful machine, one server can easily provide all of these services.




If, however, many clients are on a network and they all need a variety of services, it’s often a good idea to add one or more additional servers. With a second, third, and fourth server added, network efficiency improves. In a home, you’re not likely to need more than one server; in a business environment, however, using multiple servers may be necessary. You can use one server, for example, as a file and print server, and another as a backup server, Web server, database server, or other type of server, depending on the type of business you run and the amount and type of information you use in that business.
When more servers are added to a network, each server usually specializes in certain jobs to help improve productivity. For example, one server might contain all the data files and folders for all clients, one server might provide e-mail and Internet access, and another server might furnish shared applications.

Security
One of the most important services a server provides is authentication, or checking to make sure that a user has permission to access the network. In a small network, one server might perform authentication and other services. In a larger network with multiple servers, an authentication server might perform the job. Similar to Windows checking your username and password, an authentication server does not allow a user access to the network unless he or she types in the exact information called for and that information appears in the server’s master list. The list contains usernames, passwords, and user account information.

User account information includes a list of permissions, or rights, for a specific user on the network. These rights specify whether the user can access certain files, folders, and services. If a user wants to print, for example, the client computer sends a request to the server, which then either accepts or denies the request according to that user’s permissions. If the server accepts the request, the print job is then sent to the appropriate printer. The network administrator sets the rights and permissions for each user of the network.

Using a client/server network

A client/server network is different from a workgroup network in that only one computer called the server shares its resources with all other computers called clients. The computers in a workgroup network could also be called clients; a client is a device that makes use of the available services. In a client/server network, the clients only use the server’s services.

A server generally stores all data files, many applications, and control over peripherals and resources. Some networks contain one server, and others may contain multiple servers. The number of servers depends on the number of clients and the number of services offered.

The server controls all files, folders, printers, and other resources. The client computers must request a service and be approved before they can access a resource. Note that cables connect all devices to a small box, called a hub. Most networks need a hub to help modify transmission signals, and to extend the network past two workstations.

A client/server network, depending on the hardware, services offered, and the network operating system, can serve anywhere from 10 to 150 or so clients. As in a workgroup network, the more clients attached to the network, the busier the network traffic and perhaps the slower the transmissions. Some larger networks have multiple servers to provide a wide variety of services to the clients.