|
[ Pobierz całość w formacie PDF ]
.However, if DHCP is not available, you will want to configure TCP/IP manually.TCP/IP requires that all hosts use unique 32-bit addresses.These addresses areexpressed in the form of dotted decimal numbers, such as 192.168.1.The reasonwe use the dotted decimal format is to make it easier for us to use the number.Each segment of the dotted decimal represents 8 bits of the 32-bit number;because they use 8 bits, these segments are sometimes referred to as octets.Part of this number represents the unique host address, and another part repre-sents a network address.Computers that are connected on the same physical net-work segment use the same network address, but unique host addresses.If yourcomputer wishes to communicate with another computer that uses a differentwww.syngress.com 189_XP_06.qxd 11/12/01 9:26 AM Page 321Windows XP Networking " Chapter 6 321network address, your computer must communicate with a router that will for-ward the packets from your computer to a different network.Configuring & Implementing&Resetting TCP/IPBecause TCP/IP is a core component of the Windows XP operatingsystem, you cannot uninstall it.However, in situations that might call forthe reinstallation of TCP/IP, you can reset the protocol using the NetShellutility.Resetting TCP/IP has the same effect as uninstalling and rein-stalling the protocol by returning it to its state at the installation of theoperating system.For more information on this, see the MicrosoftKnowledge Base article Q299357 at http://support.microsoft.com.Imagine that the IP addresses represent house addresses.Each house on thestreet has a unique house number, but the complete address will use a commonstreet address for each house on the same street.If you want to deliver a letter to ahouse on the same street, you can walk the letter to the house yourself.However, ifyou want to deliver the letter to a house on a different street, you need to use theservices of the post office (router) to get your letter to the correct destination.In order to distinguish the host portion of the IP address from the networkportion of the address,TCP/IP uses the subnet mask.The subnet mask tellsTCP/IP how many bits in the address are used to represent the host and the net-work portions of the address.For example, a subnet mask that is expressed as255.255.255.0 will tell TCP/IP that the first three octets are used for the net-work portion of the address, and the last octet is used for the host portion of theaddress.Therefore, given a subnet mask of 255.255.255.0, a computer with anaddress of 172.16.17.2 and anopther with an address of 172.16.25.8 would beseen by TCP/IP to be on different networks, because the portion of the address masked by the subnet mask changes (172.16.17 and 172.16.25).However, if wewere to change the subnet mask to 255.255.0, both computers would be seenby TCP/IP to be on the same network, because the portion of the address masked by the subnet mask (172.16) does not change.Computers use binary numbers (0s and 1s).This is true of TCP/IP as wellcomputer names and dotted decimal notation are something we use to make itwww.syngress.com 189_XP_06.qxd 11/12/01 9:26 AM Page 322322 Chapter 6 " Windows XP Networkingeasier to remember addresses and work with numbers.When one computer triesto communicate with another using TCP/IP, it will AND its subnet mask withits own IP address and the IP address of the remote computer.ANDing is analo-gous to multiplication and is the process of performing a bitwise operation onbinary numbers.Any time we AND a 0 with a 1, the result is 0; any time weAND a 1 with a 1, the result is a 1.If the results of the ANDing are the same forboth addresses,TCP/IP will see both addresses as being on the same network.Ifthey are different,TCP/IP will see the addresses as being on different networks.Designing & Planning&Binary NumberingBinary numbering uses two digits, 0 and 1.Binary numbers work like allnumbering systems, including decimal.A decimal number such as 123can be expressed (1 x 102) + (2 x 101) + (3 x 100) = 100 + 20 + 3 = 123.Keep in mind that any number raised to the power of zero is one.Witha binary number, we do something similar, except we are working witha base 2, rather than a base 10, number.Therefore, a binary numbersuch as 1101 could be expressed as (1 x 23) + (1 x 22) + (0 x 21) +(1 x 20) = 8 + 4 + 0 + 1 = 13.A binary number such as 11111111 couldbe expressed as 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1 = 255.Breakingthe 32-bit TCP/IP address into four units of 8 bits each (octets) makesthem easier to work with [ Pobierz całość w formacie PDF ] |
Archiwum
|