Differentiating Among Protocols
| You might find yourself working with a number of protocols in today's networked environments. The primary function of these protocols is to facilitate communication between network devices. This section reviews the main characteristics of the most widely used protocols. Connectionless and Connection-oriented ProtocolsBefore getting into the characteristics of the various network protocols and protocol suites, it's important to first identify the difference between connection-oriented and connectionless protocols. In a connection-oriented communication, there is guaranteed delivery of the data. Any packet that is not received by the destination system is resent by the sending device. Communication between the sending and receiving devices continues until the transmission has been verified. Because of this, connection-oriented protocols have a higher overhead and place greater demands on bandwidth.
In contrast to connection-oriented communication, connectionless protocols offer only a best-effort delivery mechanism. Basically, the information is sentthere is no confirmation that the data has been received. If there is an error in the transmission, there is no mechanism to resend the data, so transmissions made with connectionless protocols are not guaranteed. Connectionless communication requires far less overhead than connection-oriented communication, so it is popular in applications such as streaming audio and video where a small number of dropped packets might not represent a significant problem.
Internetwork Packet Exchange/Sequenced Packet ExchangeLike TCP/IP and AppleTalk that are also discussed in this chapter, IPX/SPX is not a single protocol but rather a protocol suite. IPX/SPX was created by Novell for use on Novell networks. When Novell had a larger presence in the network arena, so too did the IPX/SPX protocol suite. Today, the popularity of IPX/SPX has yielded to TCP/IP although it is still used in some network environmentsenough at least to include it in the CompTIA exam objectives. TCP/IP's suitability for large multisite networks and its general acceptance has now even led Novell to adopt TCP/IP as the protocol of choice. Table 4.3 shows some of the protocols that comprise the IPX/SPX suite and their functions.
IPX AddressingAn example of an IPX address is 0BAD33CE:0003FE7C06EC. The 0BAD33CE portion represents the IPX address for the network, which is also sometimes referred to as the network number. The part 0003FE7C06EC is the MAC address of the node, which is used for the second part of the address. The node MAC address is derived directly from the MAC address burned on to each network card, but in IPX addressing, it is expressed without the colons (:). In addition to this format, IPX addresses can also be written with each group of four hexadecimal characters separated by colonsfor example, 0000:0007:003C:7F53:04CF. In some cases, any leading 0s on the network address portion are dropped. For example, 00000007 can be expressed simply as 7. The address would then be 7:003C:7F53:04CF.
IPX InteroperabilityAs you might expect, the IPX/SPX protocol suite is fully supported by Novell NetWare, but it can also be used in a Microsoft Windows environment. Microsoft includes its own version of the IPX/SPX protocol, NWLink, which provides this interoperability. Using the NWLink protocol and the Microsoft Client for NetWare, Windows systems can connect to a NetWare server using IPX/SPX. Because of the prevalence of TCP/IP, interoperability with the IPX/SPX protocol has become less important. For some time now, TCP/IP has been used as the default protocol on Novell networks. As far as Linux is concerned, there is a way to use the IPX/SPX protocol on a Linux system, but TCP/IP is the protocol of choice there too. IPX/SPX NamingUnlike TCP/IP, which is discussed later, there are few issues with IPX/SPX naming because servers are normally the only parts of a network that are assigned names. These names, which are sometimes referred to as addresses, can be up to 47 characters (in current versions of NetWare). Workstations do not need such names and instead just use IPX addresses. NetBEUI ProtocolNetBEUI was once a popular protocol for smaller networks. It is fast and easy to configure but has one significant drawback in that it is not routable. This one fact limits NetBEUI to a single network segment far too restrictive for the majority of today's networking environments. NetBEUI AddressingIn terms of addressing, NetBEUI is perhaps the simplest of all the protocols discussed here. For this reason, it is still sometimes used on very small simple networks such as those found in a home or on very small business networks. Computers on a NetBEUI network are identified by NetBIOS names. The NetBIOS name can be no longer than 15 characters and must be unique to the network. Using the 15 characters, you can assign the computers descriptive names such as workstation, student1, or secretary2.
AppleTalkAppleTalk is a protocol associated with Apple networks. The AppleTalk protocol is an established protocol, having been introduced in the early 1980s, and continued development toward the end of the 1980s enabled it to become a viable internet-working protocol. Like the IPX/SPX and TCP/IP protocol suites, the AppleTalk protocol suite is composed of several protocols. Table 4.4 lists the protocols within the AppleTalk protocol suite and their functions.
AppleTalk AddressingLike the other protocols discussed, the AppleTalk protocol uses a two-part addressing schemea node and a network section. The node portion of the address is assigned automatically when the system is first brought up onto the network. It is a randomly generated number and then broadcast to the entire network. If a duplicate node address is assigned, another will be assigned and rebroadcast to the network. The network portion of the address is assigned by the network administrator. The actual AppleTalk address is 24 bits long with 16 bits used for the network address and 8 bits for the node address. AppleTalk addresses are expressed in decimal format, with the network and node addresses separated by a period. An example of an AppleTalk address might be 4.67. The 4 represents the network number, and 67 is the node number.
AppleTalk InteroperabilityAppleTalk was designed for the purpose of being used on Apple networks and, as such, is not natively supported by most of the other major operating systems. Because of this, today, other protocols such as TCP/IP are a more common choice, even for Apple-based networks. In fact, Macintosh systems themselves support the use of TCP/IP. AppleTalk can be configured to work with other platforms, but, given the proliferation of TCP/IP, this is not widely done. AppleTalk RoutingThe earliest implementations of AppleTalk were not routable, but later versions were. Routing functionality for AppleTalk is provided by the RTMP protocol. RTMP provides similar functionality to the RIP protocol used with IPX/SPX and TCP/IP networks. AppleTalk NamingAppleTalk networks use logical hostnames, making systems readily recognizable on the network. The network address-to-hostname resolution is handled by the NBP protocol in the AppleTalk protocol suite. It performs a similar function to that provided by DNS on a TCP/IP network. The TCP/IP Protocol SuiteQuite often, TCP/IP is referred to as a network protocol, although that's not entirely accurate. Like IPX/SPX and AppleTalk, TCP/IP is actually a protocol suite comprised of many separate protocolseach of which has its own purpose and function. Combined, they all provide the TCP/IP functionality. The following list contains some of the more well-known protocols found within the TCP/IP protocol suite:
This is just an introduction to the protocols found within the TCP/IP protocol suite. Chapter 5, "TCP/IP (Transmission Control Protocol/Internet Protocol)," as well as objectives 2.5 through 2.12, discuss TCP/IP in much more detail. TCP/IP StandardsOne of the strengths of the TCP/IP protocol suite is that it is not owned by any one party and is not licensed. This is in contrast to protocols such as AppleTalk and IPX/SPX, which are owned by Apple and Novell, respectively. Because of its non-proprietary nature, TCP/IP has an open development model with its standards published in documents known as Requests for Comments (RFCs). RFCs are maintained by the Internet Engineering Task Force (IETF). You can find RFCs pertaining to TCP/IP on IETF's website at www.ietf.org. TCP/IP AddressingAnyone who has worked with TCP/IP knows that TCP/IP addressing can be a complex topic. This section provides an overview of TCP/IP addressing to compare how other protocols handle addressing. However, Chapter 5 provides a detailed look at the TCP/IP protocol including addressing. In the most commonly deployed version of TCP/IP, version 4, (IPv4) addresses are composed of four sets of 8 bits referred to as octets. These are expressed in numbers and separated by periods. An example of a TCP/IP address is 192.168.3.2. This format is often referred to as a 32-bit dotted decimal. A single TCP/IP address represents both the IP address of an individual system and the network to which the system is attached. Determining which part of the IP address belongs to the network and which belongs to the node is the responsibility of the subnet mask. If part of the address refers to the network, it is assigned a binary value of 1 within the subnet mask. If it is the node address, it's assigned a binary value of 0 within the subnet mask. For example, if you had a subnet mask of 255.255.255.0, the first two octets refer to the network and the second refer to the node address. So using the previous IP address as an example, the 192.168.3 portion of the address represents the network ID, and the .2 portion of the address represents the node ID. Table 4.5 shows default subnet masks and addressing examples.
As previously mentioned, more information on TCP/IP addressing is provided in Chapter 5. TCP/IP InteroperabilityOf all the protocols used on today's networks, TCP/IP is by far the most versatile and interoperable. All of the popular operating systems today not only support TCP/IP, but the vast majority also use it as the default protocol. This means that in any network environment, you can have Linux, Windows, and NetWare servers and clients all communicating using TCP/IP. TCP/IP NamingSystems on a TCP/IP network can be accessed from the network either by their IP address or by a hostname. Hostnames are the names assigned to the system to make them easier to remember. For instance, the secretary's computer might have the address of 192.168.4.23, but you can access it using its hostname of secretary1 or whatever name you assign it. The name-resolution process from IP address to hostname is often performed dynamically through a Domain Name Server (DNS). It can also be done statically using a text file called 'Hosts,' which is stored on each system. More information on name resolution is provided in Chapter 5. TCP/IP RoutingTCP/IP is a fully routable protocol, making it a natural choice for large networks and those that span multiple locations. As mentioned previously, TCP/IP is a protocol suite; there are two primary protocols within TCP/IP that provide the routing functionalityRouting Information Protocol (RIP) and Open Shortest Path First (OSPF). Protocol SummaryTo help you in your exam preparations, the most pertinent information from this section is listed in Table 4.6.
|
