NETWORKING PROTOCOLS in Software

Creation Code-39 in Software NETWORKING PROTOCOLS

Protocols are the standards or rules that allow many different systems and devices to interconnect and operate on a network, be it a small office network or the worldwide Internet Protocols specify how and when something should happen, and what it is that should happen Protocols are developed in many different ways, but for one to become widely used, it must be accepted by a number of organizations, most importantly those building networking software and networking devices To gain this acceptance, protocols are initially circulated as draft protocols, using documents called Requests for Comments (RFCs) If someone or some organization wants to change a protocol, a new RFC gets circulated with a higher number than the original RFC The Information Sciences Institute (ISI) of the University of Southern California classifies RFCs as being approved Internet standards, proposed Internet standards, Internet best practices, and For Your Information (FYI) ISI also maintains a web site that lists RFCs and their classification (http://www rfc-editororg/) from which you can get copies of the RFCs You can determine how a piece of software has implemented a protocol by looking at which RFCs are supported For example, you can go to Microsoft s TechNet site (http://technet2microsoftcom) and look up the topic TCP/IP RFCs and see a list of RFCs that are supported Numerous protocols deal with computer networking, but networking protocols normally refers to the logical addressing and transfer of information These protocols deal with the Transport and Network layers of the OSI networking model, as shown in Figure 5-16 While several networking protocols were originally used, today almost all networking uses Transmission Control Protocol/Internet Protocol or TCP/IP

Transmission Control Protocol/Internet Protocol is a set of networking protocols that grew out of the Internet and has been refined over 25 years to be an excellent tool for transmitting large amounts of information reliably and quickly over a complex network The two components, TCP and IP, were originally combined and later were separated to improve the efficiency of the system Until recently (around 2005), if someone said IP they automatically meant IPv4, the version of IP that had been in existence for over 25 years and provided 32-bit or 4-byte addressing While 32-bit addressing provides for a large number of addresses (over 2 billion of them), IP addressing is used worldwide, and using IPv4, the world will run

out of addresses In addition a number of technological improvements have impacted the Internet protocol Therefore in the mid-1990s discussion began on a new Internet protocol, now called IPv6, with a number of improvements including 128-bit or 16-byte addressing IPv6 has now become an official standard, and Windows Vista and Windows Server 2008 support both IPv4 and IPv6 The worldwide implementation of IPv6 will take a number of years, so IPv6 is made to work with IPv4, and systems for some time will have to handle both

The Internet Protocol operates at the Network layer, Layer 3, of the OSI model controlling the assembly and routing of packets (sometimes called datagrams in IP) To send a packet to a remote node across a complex network, such as the Internet, these are the steps that take place: 1 The packet is assembled at the Network layer, and the IP address of the destination (which is a logical address, not a physical address) is added 2 The packet is passed to the Data Link layer, which packages the packet in a frame and adds the physical address of the first router that starts the packet on its way to its destination

5:

3 The packet is passed to the closest router, which unpacks the frame, looks at the logical address on the packet, repackages the packet in a frame, and adds the physical address of the next router that continues the packet on its way to its destination 4 Step 3 repeats for each router along the path 5 At the last router before the destination, the destination physical address is added to the frame and the frame is sent to its destination This technique allows the information to follow any path through the network, with the decision on which path to take made by the local router, drawing on its knowledge of the local situation IP is a connectionless service It sends the packet on its way not knowing whether it is received or what route it took Determining whether it is received and replacing the packet if it isn t is the function of TCP at the Transport layer The route that the packet takes is a function of the routers at the Data Link layer The packet that IP assembles has a maximum of up to 64KB, including a variable header with the source and destination addresses and other control information The header, which is always a multiple of 4 bytes or 32 bits, is shown in Figure 5-17 and has these fields (each field shows the field length in bits): Version Protocol version number, set to 4 IHL Header length in terms of the number of 4-byte (32-bit) words (five-word minimum in Figure 5-17 each row is a word, so the IHL is 6)