Server-based networking provides you with the ability to build large networks that offer a greater range of resources to users (when compared to peer-to-peer networks). This is because a number of different, specialized server types (such as mail and database servers) can be included on the network.
Server-based networks also provide you with greater centralized control of resources and make it easier to add additional computers, users, and resources (again, when compared to peer-to-peer networks). Server-based networks are scalable networks , meaning they are easily expandable.
One requirement for a server-based network is a computer running a network operating system; this computer is known as the server . As already mentioned in Chapter 1, "Computer Networking Overview," a server computer is basically a special-purpose machine that logs in users and "serves" up resources, such as files and printers, to the users. Because the server verifies users and determines the level of access the users will have to the various network resources, server-based networks provide a more secure environment than peer-to-peer networks.
Actually accessing resources on a server-based network is also easier than in the peer-to-peer environment because one username and one password gets a user onto the network and provides that user access to any resource he or she has the appropriate permissions for. This is in sharp contrast to a peer-to-peer network, which may have a different password for every resource on the network.
Pros and Cons of Server-Based Networking
As is the case with peer-to-peer networking, server-based networking has its pros and cons. The upside of server-based networking revolves around the fact that this type of network provides central control of resources and makes it easier for users to actually find resources. For example, the network operating system (NOS) Microsoft Windows Server 2003 manages resources such as shared folders and drives , printers, and even users in a tool called the Active Directory, shown in Figure 2.2 (this is how the Active Directory tool would appear to a network administrator).
Figure 2.2. Network operating systems such as Windows Server 2003 provide for the central management of users and resources.
Active Directory is used to add and remove users from the network and can even be used to place users who access the same resources into groups. Management tools such as Active Directory provide the administrator with the ability to control network access and the different levels of access that are given to users or groups of users. Information on configuring user accounts for network access is discussed in more detail in Chapter 20, "A Network Security Primer," in the section "Administrators and Users."
This ability to manage network users and resources centrally comes with a high price tag, however, and one of the major cons of server-based networking is the cost of the dedicated server computer and the NOS that you must run on it. Server-based networks also require a network administratorsomeone who is well versed in the NOS being used. This usually means an additional employee on the company's payroll, which is another cost associated with server-based networks.
However, the overall cost of network operating system software and computer hardware is lower today than it ever has been, and the server-based network has become the standard for networking in even relatively small companies. Security features built in to the NOS allow the network administrator to protect the company's data from outside the network, and they also provide a great deal of control over sensitive data and its access from inside the network.
Let's break the pros and cons of server-based networks into two lists, with the pros first:
Now let's look at the cons:
The local area network (LAN)another name for a server-based network in one locationis really the rule rather than the exception in today's business world. Even small LANs can use different types of servers to provide users with the resources they need. Let's take a look at the possibilities.
Types of Network Servers
We've already discussed the fact that a LAN uses a server to validate users as they log on to the network. If a user's login name and password don't match up, the server doesn't let him or her on the network. In large networks, this "central computer" (for lack of a better name) can be kept quite busy logging users on to the network as they fire up their client PCs. Rather than overburden this server with additional duties , it's not uncommon to distribute the workload among other specialized servers.
A file server's job is to serve as a home for the files that are needed by users on the network. This can include files that a number of users share. These files are typically held in what is called a public folder , which can include private folders that are specific for a particular user. The great thing about using a file server is that important files reside in one place, making it very easy to back up the data periodically. The downside is that if the file server goes down, users can't get at their files.
A print server is used to host a network printer. It is basically the control conduit for the printer. Because print jobs need to be spooled (placed on the computer before they are sent to the printer) before they are printed, the print server supplies the hard drive space needed. The print server also queues up all the print jobs being directed to the printer. The network administrator can also delete print jobs and change the queue order of print jobs by accessing the print server.
A communication server runs specialized software that allows users on the network to communicate. It provides services such as electronic mail and discussion groups to allow users to share information. Two of the most popular communication packages for a LAN (and each need to be set up on a server on a LAN) are Microsoft Exchange and Lotus Notes (both of these communication packages are discussed in Chapter 11, "Working with Applications on the Network").
Application servers host various applications such as specialized databases. Even typical desktop applications such as word processors and spreadsheet software can also be stored on an application server. This makes updating software applications much easier because the software doesn't actually reside on every client workstation; users start these applications from their local computers, but the application software is actually stored on the server.
Web servers provide you with the ability to create a Web site that can be accessed internally by your employees (this is called an intranet ) or by folks surfing the Internet. Web servers aren't for everyone, and many companies still use Web-hosting companies to get their Web sites up and running on the Internet. A number of different software packages can be used to set up a Web server, and they vary in ease of use and stability. Microsoft Windows Server 2003 ships with Internet Information Server 6.0 (IIS6), a Web server software package. Figure 2.3 shows the IIS6 management console that is used to configure a Web server on a Microsoft network.
Figure 2.3. Web servers are used to host corporate or personal Web sites.
Other Specialized Servers
There are also other types of servers that can reside on the network that don't actually provide any resources to the LAN users, but they are needed to make the network run properly or more efficiently . For example, on larger networks that use the TCP/IP network protocol, each computer must be configured with a unique Internet Protocol (IP) address (which is much like a person's Social Security number) to communicate on the network (IP addressing is discussed in detail in Chapter 5 in the section "TCP/IP"). This means that the network administrator might have to configure hundreds of computers, one at a time. However, a server running the Dynamic Host Configuration Protocol (DHCP) can actually automatically assign IP addresses to computers as they are started up on the network.
DHCP can be run on your main network server or other server on the network. Windows Server 2003, various Linux distributions, and Novell NetWare can provide services such as DHCP on the network. Server has a built-in version of DHCP. This means a server, such as one running Windows Server 2003, can be configured to provide IP addresses (this range of addresses is called a scope ) to the computers on the network dynamically. This saves you a lot of time having to statically configure the IP address for each and every computer on the network. Figure 2.4 shows the DHCP snap-in that is used on a Windows Server to manage IP address leases.
Figure 2.4. DHCP can be used to provide IP addresses to network clients .
Other specialized servers you might need on a LAN relate to the fact that operating systems all assign some kind of friendly name to a computer. For example, think about the World Wide Web; when you want to go to a Web site, you type the Web site's name (also known as a Uniform Resource Locator or URL ) into your Web browser. You don't typically enter the actual numerical address of the site (which is an IP address). Special servers on the Internet, called DNS servers (they run the Domain Name Service), are used to resolve that friendly name you entered into an actual address. It is not uncommon to also set up DNS servers on LANs to provide this same type of service.
You've probably noticed that the needs of your users and the size of your network dictate how many of these specialized servers you have to set up on your LAN. The larger your LAN is, the greater the need to distribute the day-to-day workload among different servers on the network. In contrast, a small network might use just one server that logs on users and supplies printer and file services.