
Network Plus 2005 In Depth
.pdf
412 Chapter 8 NOS AND WINDOWS SERVER 2003-BASED NETWORKING
universal group—A group on a Windows Server 2003 network that allows members from one domain to access resources in multiple domains and forests.
UPN (user principal name)—The preferred Active Directory naming convention for objects when used in informal situations. This name looks like a familiar Internet address, including the positioning of the domain name after the @ sign. UPNs are typically used for e-mail and related Internet services.
UPN (user principal name) suffix—The portion of a universal principal name (in Windows Server 2003 Active Directory’s naming conventions) that follows the @ sign.
user principal name—See UPN.
virtual memory—The memory that is logically carved out of space on the hard disk and added to physical memory (RAM).
Windows Services for UNIX—A suite of applications designed to integrate Windows Server 2003 servers with UNIX-type of servers and clients. One application in this suite enables Windows and UNIX-type of servers to share directory information (when the UNIX-type of server has a directory installed). Another application enables UNIX-type of clients to view resources on a Windows Server 2003 server as if they were resources on a UNIX-type of server.
workgroup—A group of interconnected computers that share each others’ resources without relying on a central file server.
Review Questions
1.To better organize and manage objects, a network administrator places objects in
_________________________.
a.classes
b.objects
c.organizational units
d.attributes
2._________________________ is the ability of a processor to perform many different operations in a very brief period of time.
a.Multitasking
b.Paging
c.Formatting
d.Replication

REVIEW QUESTIONS |
Chapter 8 413 |
3.A _________________________ is a routine of sequential instructions that runs until it has achieved its goal.
a.schema
b.file system
c.namespace
d.process
4._________________________ is the original PC file system that was designed in the 1970s to support floppy disks and, later, hard disks.
a.NTFS
b.FAT
c.CDFS
d.DOS
5._________________________ is the name that uniquely identifies an object within a container.
a.Schema
b.Relative distinguished name
c.Domain name
d.Common name
6.True or false? A redirector belongs to the Presentation layer or the OSI Model.
7.True or false? A FAT16 partition or file cannot exceed 2 GB.
8.True or false? FAT is secure, reliable, and makes it possible to compress files so that they take up less space.
9.True or false? NTFS keeps a log file of system activity to facilitate recovery if a system crash occurs.
10.True or false? In Internet terminology, domain model refers to the complete database of hierarchical names used to map IP addresses to their hosts’ names.
11._________________________ involves associating a letter, such as M: or T:, with a disk, directory, or other resource.
12.The _________________________ account is a predefined user account that has the most extensive privileges for resources both on the computer and on the domain that it controls.

414Chapter 8 NOS AND WINDOWS SERVER 2003-BASED NETWORKING
13.The term _________________________ refers to the RAM chips that are installed on the computer’s system board and whose sole function is to provide memory to that machine.
14.The _________________________ lists all computer components proven to be compatible with Windows Server 2003.
15.The relationship between two domains in which one allows the other to authenticate its users is known as a(n) _________________________.

Chapter 9
Networking with
UNIX-Type of
Operating Systems
After reading this chapter and completing the exercises, you will be able to:
■Describe the origins and history of the UNIX operating system
■Identify similarities and differences between popular implementations of UNIX
■Describe the features and capabilities of servers running Solaris, Linux, and Mac OS X Server
■Explain and execute essential UNIX commands
■Install and configure Linux on an Intel-based PC
■Manage users, groups, and file access permissions in Solaris, Linux, and Mac OS X Server
■Explain how computers running other operating systems can connect to UNIX servers

Along with Microsoft Windows and Novell NetWare, UNIX is one of the most popular network operating systems. All of these operating systems enable servers to provide resource sharing, but UNIX differs in fundamental ways from NetWare and Windows. Researchers at AT&T Bell Laboratories developed UNIX in 1969; thus, it is much older than NetWare and Windows. In fact, UNIX preceded and led to the development of the TCP/IP protocol suite in the early 1970s. Today, most Internet servers run UNIX. Reflecting this operating system’s efficiency and flexibility, the number of installed UNIX-type of systems continues to grow. Many local and wide area networks include UNIX servers. You should familiarize yourself with
UNIX so you can set up and maintain these networks.
Mastering UNIX can be difficult because, unlike Windows and NetWare, it is not controlled and distributed by a single software manufacturer. Instead, numerous vendors sell their own UNIX varieties. In addition, operating systems that share many of UNIX’s characteristics but are nonproprietary and freely distributed are available. The most popular example is Linux. Fortunately, the differences between UNIX varieties are relatively minor, and, if you understand how to use one, with a little effort you can understand how to use another. This chapter introduces the UNIX operating system in general and then describes three varieties—Solaris, Mac OS X Server, and Linux—in more detail.
A Brief History of UNIX
In the late 1960s, a few programmers grew dissatisfied with the existing programming environments. In particular, they didn’t like the cumbersome nature of systems that required a programmer to write a set of instructions, submit them all at once, and then wait for the results. Instead, programmers desired a more interactive operating environment that allowed them to build and test their programs piece by piece. In addition, the programmers sought a system that imposed as few predetermined structures as possible on the users. For example, they preferred to let programmers decide how data should be formatted within a data file or how to structure filenames. With this in mind, Ken Thompson and Dennis Ritchie, two employees at Bell Labs (which was then part of AT&T) in Murray Hill, New Jersey, decided to create an entirely new programming environment. To properly design this new environment, they decided to start at the lowest level—the operating system. This environment ultimately evolved into the UNIX operating system.
Antitrust law prohibited AT&T from profiting from the sale of computers and software during the 1970s. Thus, for a nominal licensing fee, anyone could purchase the source code—that is, the form of a computer program that can be read by people—to the work produced at Bell Labs. The word spread rapidly, and researchers in educational institutions and large corporations all over the world soon had this new software running on their lab computers. Versions of UNIX that come from Bell Labs are known as System V. Researchers at the University of

VARIETIES OF UNIX |
Chapter 9 417 |
California at Berkeley were among the first enthusiastic supporters of early versions of UNIX. They added many useful features to the operating system, including the TCP/IP network subsystem. Berkeley versions of UNIX are known as BSD (Berkeley Software Distribution).
The 1980s saw the breakup of AT&T and the removal of some of its antitrust restrictions. This enabled the company to begin marketing the UNIX system to other computer manufacturers. However, AT&T eventually sold its rights to the UNIX system, and these rights changed hands a number of times during the early 1990s. Today, ownership of the UNIX system is shared by two organizations—The SCO Group and The Open Group.
The SCO Group is the company that owns the rights to the UNIX source code, or the raw materials for creating a UNIX system. Anyone could try writing a UNIX operating system from scratch, but the effort required to do so is prohibitive. Most organizations choose to start with the existing source code by licensing it from The SCO Group and then make modifications for their specific computer hardware.
In addition, a nonprofit industry association called The Open Group owns the UNIX trademark. After a vendor changes the code licensed from The SCO Group, its modified system must pass The Open Group’s verification tests before the operating system may be called UNIX. IBM, for example, pays source code licensing fees to The SCO Group and verification and trademark use fees to The Open Group so that it can refer to its AIX operating system as UNIX.
Now that you understand some of the history of UNIX, you are better prepared to learn about its many varieties and how these varieties are related.
Varieties of UNIX
NET+ |
Today, the UNIX operating system comes in many varieties, or, as they are more casually called, |
3.1flavors. Before learning about their differences, however, you should know that all flavors of the UNIX operating system share the following features:
The ability to support multiple, simultaneously logged-on users
The ability to coordinate multiple, simultaneously running tasks (or programs)
The ability to mount—or to make available—disk partitions upon demand
The ability to apply permissions for file and directory access and modification
A uniform method of issuing data to or receiving data from hardware devices, files, and running programs
The ability to start a program without interfering with a currently running program
Hundreds of subsystems, including dozens of programming languages
Source code portability, or the ability to extract code from one UNIX system and use it on another
Window interfaces that the user can configure, the most popular of which is the X
Window system

418 |
|
|
Chapter 9 NETWORKING WITH UNIX-TYPE OF OPERATING SYSTEMS |
|
|
||||
|
|
|
|
|
|
|
Types of the UNIX operating system can be divided into two main categories: proprietary and |
||
NET+ |
|
|
3.1open source. The following sections describe characteristics of each category and offer examples of UNIX versions in both categories. Note that some flavors of UNIX will operate only on certain types of computers. Examples of such UNIX varieties and their unique hardware requirements are also described next.
Proprietary UNIX
Many companies market both hardware and software based on the UNIX operating system. An implementation of UNIX for which the source code is either unavailable or available only by purchasing a licensed copy from The SCO Group (costing as much as millions of dollars) is known as proprietary UNIX. By most counts, the three most popular vendors of proprietary UNIX are Apple Computer, Sun Microsystems, and IBM. Apple’s proprietary version of UNIX, Mac OS X Server, runs on its PowerPC-based computers. Sun’s proprietary version of UNIX, called Solaris, runs on the company’s proprietary SPARC-based (the CPU invented by Sun Microsystems) workstations and servers, as well as Intel-based Pentium-class workstations and servers. IBM’s proprietary version, AIX, runs on its PowerPC-based computers. Apple and IBM use the same central processing unit, the PowerPC, in their computers. But other hardware in their computers is very different, and the software that runs on IBM computers will not run on Apple’s computers. Many other organizations have licensed the UNIX source code and created proprietary UNIX versions that run on highly customized computers (that is, computers that are appropriate for very specific tasks).
Choosing a proprietary UNIX system has several advantages:
Accountability and support—An organization might choose a proprietary UNIX system so that when something doesn’t work as expected, it has a resource on which to call for assistance.
Optimization of hardware and software—Workstation vendors who include proprietary UNIX with the computers they sell invest a great deal of time in ensuring that their software runs as well and as fast as possible on their hardware.
Predictability and compatibility—Purveyors of proprietary UNIX systems strive to maintain backward-compatibility with new releases. They schedule new releases at somewhat regular, predictable intervals. Customers usually know when and how things will change with proprietary UNIX systems.
One drawback of choosing a proprietary UNIX system, however, relates to the fact that the customer has no access to the system’s source code and, thus, cannot customize the operating system. Open source UNIX solves this problem.
Open Source UNIX
An interesting factor in the UNIX marketplace in recent years has been the emergence of UNIX-type of systems that are not owned by any one company. This software is developed and packaged by a few individuals and made available to anyone, without licensing fees. This

THREE FLAVORS OF UNIX |
Chapter 9 419 |
NET+ |
would be equivalent, in the Windows world, to a large group of people getting together and |
3.1writing a version of Windows based on the public specifications. (UNIX, however, has been publicly available for much longer than Windows, and Microsoft keeps the specifications to many parts of Windows a closely guarded secret.)
Often referred to as open source software, or freely distributable software, this category includes UNIX-type of systems such as GNU (an acronym that stands for GNU’s Not UNIX), BSD, and Linux. These systems, in turn, come in a variety of implementations, each of which incorporates slightly different features and capabilities. As mentioned, these packages are often referred to as the different flavors of the open source software. For example, the different flavors of Linux include Red Hat, Fedora Core, SUSE, Mandrake, and a host of others. Although these packages are available at no cost, it’s also possible to purchase the software, thereby gaining a certain amount of convenience. In the case of purchasing Red Hat Enterprise Linux, for example, you are paying for the convenience of a package that includes the software on CD-ROM, documentation, and access to Red Hat’s customer support. Choosing to obtain Fedora Core Linux for free from Red Hat means investing a fair amount of time, enough to download more than 2 GB of data, write the software to a DVD-ROM, and learn how to install the software without the aid of printed documentation. Many people find the convenience worth the nominal purchase price of software with a licensing fee.
The key difference between freely distributable UNIX-type of systems and proprietary implementations of UNIX relates to the software license. Proprietary UNIX includes agreements that require payment of royalties for each system sold and that forbid redistribution of the source code. A primary advantage of open source UNIX and Linux is that users can modify their code and thereby add functionality not provided by a proprietary version of UNIX. For example, a manufacturing company that uses computer-controlled robotic spot welders might combine open source UNIX or Linux with custom software to control its robots. In contrast, it might be very difficult or costly to integrate the robotic control software with a proprietary UNIX system.
Another potential advantage of using a freely distributable UNIX-type of system is that these varieties run not only on Intel-based processors, but also on other processor brands, such as the PowerPC used in Apple Macintoshes, the PowerPC used in IBM servers, and SPARC, used in Sun Microsystems workstations. Proprietary UNIX systems cannot be installed on such a wide range of systems.
Three Flavors of UNIX
NET+ |
The rest of this chapter focuses on three UNIX or UNIX-type of operating systems: Solaris, |
3.1Linux, and Mac OS X Server. The following list summarizes some basic information about these three varieties:
Solaris, the UNIX system Sun Microsystems uses on its SPARC-based servers, offers users all the benefits of commercially supported operating systems. It has seen numerous revisions and improvements over the years, and now runs behind the scenes of some of the most intensive applications in the world. Some examples

420 Chapter 9 NETWORKING WITH UNIX-TYPE OF OPERATING SYSTEMS
NET+
3.1
include large multiterabyte databases, weather prediction systems, and large economic modeling applications.
Linux follows standard UNIX conventions, is highly stable, and is free. Linus Torvalds developed it in 1991 when he was a second-year computer science student in Finland. After developing Linux, Torvalds posted it on the Internet and recruited other UNIX aficionados and programmers to help enhance it. Today, Linux is used for file, print, and Web servers across the globe. Its popularity has even convinced large corporations that own proprietary UNIX versions, such as IBM, HewlettPackard, and Sun Microsystems, to publicly embrace and support Linux.
Mac OS X Server is a relative newcomer to the UNIX server marketplace. Apple Computer released Mac OS X to the public in the Spring of 2001 to replace its previously released network operating system software, AppleShare IP. It includes the traditional Apple Macintosh user interface. Mac OS X Server runs on Apple’s Xserve line of computers as well as the Power Mac computers. These PowerPCbased servers are used by marketing organizations as well as others requiring high performance hardware and the familiar Macintosh user interface.
All UNIX-type of systems offer a host of features, including the TCP/IP protocol suite and all applications necessary to support the networking infrastructure as a part of the basic operating system. UNIX-type of systems also support non-IP protocols, such as Novell’s IPX/SPX and AppleTalk (the original Mac OS network communication protocol). In addition, when you use a UNIX-type of system, you get the programs necessary for routing, firewall protection, DNS services, and DHCP services. Like Windows Server 2003 and NetWare, UNIX also supports many different network topologies and physical media, including Ethernet, Token Ring, FDDI, and wireless LANs. And as described later in this chapter, UNIX-type of systems can also run applications and utilities that make them compatible with other NOSs on the same network.
UNIX-type of systems efficiently and securely handle the growth, change, and stability requirements of today’s diverse networks. The source code on which UNIX-type of systems are based is mature, as it has been used and thoroughly debugged by thousands of developers for many years. As with other modern NOSs, UNIX-type of systems allow you to change the server’s configuration—for example, assigning a different IP address to an interface—without restarting the server. Similarly, you can easily modify a UNIX-type of system while it is running. When you need to access a backup tape drive, for example, you can enable the tape driver, access the tape, and disable the tape driver without restarting the server. This functionality allows you to use memory on your server very efficiently.
UNIX Server Hardware Requirements
NET+ |
Hardware requirements for UNIX-type of systems are very similar to those for Windows Server |
3.12003 servers. One key difference, however, is that any UNIX-type of operating system can act as a workstation or server operating system (whereas Microsoft sells different operating sys-

UNIX SERVER HARDWARE REQUIREMENTS |
Chapter 9 421 |
NET+ tems for workstations and servers). Therefore, the UNIX computer’s minimum hardware
3.1requirements depend partly on which version of UNIX-type of operating system you are installing and partly on how you intend to use the system.
A further difference is that in all flavors of UNIX, the use of a GUI (graphical user interface) remains optional—that is, you can choose to use the GUI, a command-line interface, or a combination of the two. By contrast, in Windows Server 2003, you must use the GUI for many operations (and so a mouse is required). Many people regard the choice of using a GUI or a command-line interface in UNIX as an advantage. For example, you might choose to use the GUI for operations that require a great deal of interaction, such as adding new users or configuring services. However, for server operations that run unattended, it often makes sense to use the command-line interface (which consumes less of the computer’s memory and other resources).
As you learned in Chapter 8, no single “right” server configuration exists. You’ll need to add more memory and more disk space according to your networking environment and users’ needs. Unfortunately, you sometimes cannot learn the memory requirements of an application until you actually run it on the server. In these instances, it is always better to overestimate your needs than to underestimate them. However, you can get an idea of what additional hardware your server may require by answering the following questions:
How will the server be used (for example, for file and print services or for backups)?
Which applications and services will run on the server?
How many users will this system serve?
Will all users be accessing the server during the same time periods?
What is the maximum amount of downtime the users of this server can tolerate?
The following sections provide rough guidelines for choosing the hardware you will need to run Solaris, Linux, and Mac OS X Server.
Solaris Hardware Requirements
Solaris runs on computers containing Sun SPARC processors or on Intel-based processors. Table 9-1 lists the minimum hardware requirements for the most recent Solaris operating system release, version 10. Components on SPARC workstations and servers have been tested to ensure their compatibility with Solaris. To determine which components on an Intel-based system will be compatible with Solaris, refer to Sun Microsystems’ Intel hardware compatibility list at www.sun.com/software/solaris/specs.html.
The hardware requirements listed in Table 9-1 reflect the minimum amount of memory and hard disk space required to run Solaris 10. However, for better performance, you should consider using more than the minimum requirement. For example, for application support, most systems need at least 10 GB of hard disk space.