Info Networking / Operating System / Server

UNIX & LINUX

UNIX and Linux were designed to be extremely flexible and customizable.

As a result, UNIX and Linux support dozens of user interfaces.
The most common are the text-based interfaces called shells.
Users type commands that are interpreted by the shell, which in turn relays the user instructions to the operating system and other programs. UNIX (and Linux) shells are difficult to learn because they rely on abbreviations and complex command syntax. Commonly used shells include the following:
• Bourne shell
• Korn shell
• Bash shell
• C shell
• TC shell
Enter the UNIX command, uname, on most systems to find out what version of the UNIX or Linux a computer is running.

In contrast to early desktop PC software, NOSs provide built-in networking components and network services, multiuser capability, and sophisticated file security and file sharing technologies.
NOSs must have a robust kernel to prevent crashes and downtime.
It is especially important that the NOS kernel schedule and manage multiple processes so that each program is prevented from crashing other parts of the system.
Because specialized administrators manage NOSs, they do not necessarily require resource-consuming graphical interfaces.
Finally, a NOS requires a sophisticated file system that allows for efficient storage and maximum security. Instead of FAT, a NOS typically employs NTFS, UFS (UNIX file system), or another equally robust file system.

Common NOSs in use today include:
• Microsoft Windows – Network operating systems offered by Windows are NT 3.51, NT 4.0, 2000, XP, and .NET. Keep in mind that NT 3.51 and NT 4.0 are being phased out as a result of the superior Windows 2000, XP, and .Net. Figure and Figure show Windows XP and 2000 operating system.
• Novell NetWare – Novell line of operating systems, which included NetWare 3.12, IntraNetWare 4.11, NetWare 5.0 and 5.1. Figure and Figure show examples of Novell operating system.
• Linux – Linux operating systems including Red Hat, Caldera, SuSE, Debian, and Slackware. Figure and Figure show Linux operating systems from Red Hat, and Caldera.
• UNIX – Various operating systems offered UNIX, such as HP-UX, Sun Solaris, BSD, SCO, and AIX. Figure shows Sun Solaris.
The TCP/IP suite of protocols has become the dominant standard for internetworking. It was originally defined by researchers in the United States Department of Defense. TCP/IP represents a set of public standards that specifies how packets of information are exchanged between computers over one or more networks.
Application Protocols
The following protocols function at the application layer of the OSI model:
• Telnet – Telnet enables terminal access to local or remote systems. The telnet application is used to access remote devices for configuration, control, and troubleshooting.
• File Transfer Protocol (FTP) – FTP is an application that provides services for file transfer and manipulation. FTP uses the Session layer to allow multiple simultaneous connections to remote file systems.
• Simple Mail Transport Protocol (SMTP) – SMTP provides messaging services over TCP/IP and supports most Internet e-mail programs.
• Domain Name System (DNS) – DNS provides access to name servers where network names are translated to the addresses used by Layer 3 network protocols. DNS greatly simplifies network usage by end users.
Transport Protocols
The following protocols function at the transport layer of the OSI model:
• Transmission Control Protocol (TCP) – TCP is the primary Internet protocol for reliable delivery of data. TCP includes facilities for end-to-end connection establishment, error detection and recovery, and metering the rate of flow of data into the network. Many standard applications, such as e-mail, web browsing, file transfer and Telnet, depend on the services of TCP. TCP identifies the application using it by a "port" number.
• User Datagram Protocol (UDP) – UDP offers a connectionless service to applications that do not want the overhead of TCP and can tolerate a level of data loss. Applications in network management, network file system, and simple file transport use UDP. Like TCP, UDP identifies applications by port number.
Internet Protocols
The following protocols function at the network layer of the OSI model:
• Internet Protocol (IP) – IP provides source and destination addressing and, in conjunction with routing protocols, packet forwarding from one network to another toward a destination.
• Internet Control Message Protocol (ICMP) – ICMP is used for network testing and troubleshooting. It enables diagnostic and error messages. ICMP "echo" messages are used by the PING application to test remote devices.
• Routing Information Protocol (RIP) – RIP operates between router devices to discover paths between networks. In an Internetwork, routers depend on a routing protocol to build and maintain information about how to forward packets toward their destination. RIP chooses routes based on their distance, or hop count.
• Open Shortest Path First (OSPF) – OSPF, like RIP, enables routers to build forwarding tables. Unlike RIP, OSPF selects routes based on other characteristics of the links between networks, such as bandwidth and delay. OSPF is more suitable than RIP for routing in large internetworks.
• Address Resolution Protocol (ARP) – ARP is used to discover the local (MAC) address of a station on the network when its IP address is known. End stations as well as routers use ARP to discover local addresses.
• The Ethernet architecture is the most popular type of LAN link used today. It is based on the 802.3 standard. This specifies that a network that implements the Carrier Sense Multiple Access/Collision Detection (CSMA/CD) access control method must use a baseband transmission over coaxial or twisted-pair cable that is laid out in a bus topology (that is, a linear or star bus). CSMA/CD means that multiple stations will have access to the media, and before one station can access that media, it must first listen (carrier sense) to detect if another system is using the same media. If another system is using the media, then that system must wait before it can transmit. If both systems attempt to send data at the same time, then a collision will result.
• Standard transfer rates are 10 megabit per second (Mbps) or 100 Mbps. The new standards provide for Gigabit Ethernet, capable of attaining speeds up to one Gbps over fiber-optic cable or other high-speed media. Figure shows a table of main Ethernet specifications.

10BASE-T
Currently, 10BASE-T is one of the most popular Ethernet implementations. It uses a star bus topology.
The term Ethernet cable can be used to describe the unshielded twisted-pair (UTP) cabling generally used in this architecture. Shielded Twisted-Pair (STP) also can be used. 10BASE-T and 100BASE-X create networks that are easy to set up and expand.
One advantage of 10BASE-T is that it is relatively inexpensive. Although a hub is required when connecting more than two computers, small hubs are available at a low cost, and 10BASE-T network cards are inexpensive and widely available.

NOTE:
Inside the Ethernet hub, the signaling system is a bus, as with coax Ethernet networks.
NOTE:
10BASE-T specifications require a hub. However, if connecting only two computers (for example, for a home network), and UTP is preferred rather than thinnet, a crossover cable can be used. This type of cable has wire pairs that are cross connected. Crossover cables are also used to connect two hubs to each other if the hubs do not have uplink ports.
UTP, which is the most commonly used twisted-pair cabling, is thin, flexible, and easier to work with than coax. It uses modular RJ-45 plugs and jacks, so it is easy to connect the cable to the NIC or hub.
The disadvantages of 10BASE-T are that the maximum length for a 10BASE-T segment (without repeaters) is only 100 meters (about 328 feet). Also, the UTP used is more vulnerable to electromagnetic interference (EMI) and attenuation than other cable types.

100BASE-X
100BASE-X comes in several different variations. It can be implemented over Category 5 UTP (100BASE-T), over 2-pair Category 5 UTP or STP (100BASE-TX) or as Ethernet over 2-strand fiber-optic cable (100BASE-FX).
• Advantages of 100BASE-X – Regardless of the implementation, the big advantage of 100BASE-X is high-speed performance. At 100 Mbps, transfer rates are 10 times that of 10BASE-T.
Because it uses twisted-pair cabling, 100BASE-X also shares the same advantages as 10BASE-T. These include low cost, flexibility, and ease of implementation and expansion.
• Disadvantages of 100BASE-X – 100BASE-X shares the disadvantages of 10BASE-T, which are inherent to twisted-pair cabling, such as susceptibility to EMI and attenuation. 100-Mbps NICs and hubs are generally somewhat more expensive than those designed for 10-Mbps networks, but prices have dropped as 100BASE-X has gained in popularity. Fiber-optic cable remains an expensive cabling option, not so much because of the cost of the cable itself, but because of the training and expertise required to install it.
Digital Subscriber Line (DSL) is an always-on technology. This means that there is no need to dial up each time when connecting to the Internet. It is a relatively new technology currently being offered by phone companies as an add-on service over existing copper wire or phone lines.
DSL comes in several varieties:
• Asymmetric DSL (ADSL) currently is the most common implementation. It has speeds that vary from 384 kbps to more than 6 Mbps downstream. The upstream speed is typically lower.
• High Data Rate DSL (HDSL) provides bandwidth of 768 kbps in both directions.
• Symmetric DSL (SDSL) provides the same speed, up to 3 Mbps, for uploads and downloads.
• Very High Data Rate DSL (VDSL) is capable of bandwidths between 13 Mbps to 52 Mbps.
• ISDN DSL (IDSL) has a top speed of 144 kbps but is available in areas that do not qualify for other DSL implementations. IDSL is actually DSL over ISDN lines.

In a TCP/IP-based LAN, PCs use Internet Protocol (IP) addresses to identify and locate each other. An IP address is a 32-bit binary number. This binary number is divided into 4 groups of 8 bits known as octets, each of which is represented by a decimal number in the range of 0 to 255. The octets are separated by decimal points. An example of an IP address is 190.100.5.54. This type of address is described as a dotted decimal representation. Each device on the network that has an IP address is known as a host or node.
A secondary dotted decimal number, known as the subnet mask, always accompanies an IP address. A subnet mask is a tool used by a system administrator to segment the network address that has been assigned to the network the technique of subnetting allows the entire network to be represented to the Internet by one address. An example of a subnet mask is 255.255.0.0. The subnet mask is also used to determine whether a particular host IP address is local (on the same network segment) or remote (on another segment).
There are several options for assigning IP addresses for hosts on a LAN:
• Static – Assigned by the network administrator manually
• Dynamic – Assigned by a Dynamic Host Configuration Protocol (DHCP) server (DHCP servers are discussed in the next section.)
• Automatic – Private IP addressing
If there are more than a few computers, manually configuring TCP/IP addresses for every host on the network can be a time consuming process. This also requires that the network administrator assigning the addresses understands IP addressing and knows how to choose a valid address for the particular network. An IP address is unique for each host. The IP address is stored in the Network Settings of the Operating System Software. It is commonly referred to as the logical address. TCP/IP addressing will be taught later in this course.
In the Windows operating system, the IP address is manually entered into the TCP/IP properties dialog box. Figure shows the TCP/IP configuration box, which is used to set the address settings, or configurations that are entered, which include the following:
• An IP address
• A subnet mask
• Default gateway address
• Optional values including a Domain Name System (DNS) server address and Windows Internet Naming Service (WINS)
The default gateway address and the DNS are discussed in a later section.


DHCP Server

Another way for computers on a network to obtain an IP address is through a Dynamic Host Configuration Protocol (DHCP) server. DHCP is a software utility that automatically assigns IP addresses to PCs. The computer running the software is known as a DHCP server. DHCP servers assign the IP addresses and TCP/IP configuration information to computers configured as DHCP clients. This dynamic process eliminates the need for manual IP address assignments. However, any devices requiring a static or permanent IP address must still have its IP address manually assigned. Figure shows an example of the tool, which a user would use to configure a server to run DHCP services to client systems on the network.
When the DHCP server receives a request from a host, it selects IP address information from a set of predefined addresses that are stored in its database. Once it has selected the IP information, it offers these values to the requesting device on the network. If the device accepts the offer, the DHCP server will then lease the IP information to the device for a specific period of time.
The IP address information that a DHCP server can assign to hosts that are starting up on the network includes the following:
• An IP address
• A subnet mask
• Default gateway
• Optional values including a DNS server address and WINS
The use of this system simplifies the administration of a network because the software keeps track of IP addresses. Automatically configuring TCP/IP also reduces the possibility of assigning duplicate IP addresses or invalid IP addresses. For any computer on the network to take advantage of the services provided by the DHCP server, it must first be able to identify the server on the local network. The option to obtain an IP address automatically is selected on the TCP/IP Properties dialog box. Figure shows an example of the IP address configuration screen of a Windows client when it is configured with an IP address via DHCP. In other cases, an operating system feature called Automatic Private IP Addressing (APIPA) enables a computer to assign itself an address if it is unable to contact a DHCP server.



The Domain Name System (DNS) is used to translate the computer names such as www.cisco.com to the corresponding unique IP address. The name resolution process is demonstrated in Figure . The DNS software runs on a computer acting as a network server and it makes the address translations. DNS software may be hosted on the network by itself or by an Internet service provider (ISP). Address translations are used each time the Internet is accessed. The process of translating names to addresses is known as name resolution.



3.1.5 Default gateway

A computer located on one network segment that is trying to talk to another computer on a different segment across a router, sends the data through a default gateway. The default gateway is the near-side interface of the router. It is the interface on the router to which the local computer network segment or wire is attached. In order for each computer to recognize its default gateway, the corresponding near side router interface IP address has to be entered into the host TCP/IP Properties dialog box. Figure demonstrates how the default gateway would be set up and its relationship to the other router interfaces on the network.



Networking media can be defined simply as the means by which signals, or data, are sent from one computer to another. The signals can be transmitted through cable or wireless means. There are a wide variety of networking media in the marketplace. The following media types will be discussed:
• Copper – including coaxial and twisted pair
• Glass – fiber optic
• Waves – wireless