TCPIP Operation

TCP/IP Operation

TCP/IP is a connectionless protocol, transferring information in datagrams called packets. Each of these packets is individually forwarded through the network. There are limitations to the size of the packet, and if a packet exceeding this maximum size is presented to the network, it is a function of TCP to break the data down into a more manageable stream for the network.

For example, suppose a 15,000-octet file is to be sent; most networks can't handle a 15,000-octet packet. So TCP will break this single large packet into several smaller pieces; maybe 30 500-octet packets. Each of these packets is sent to the destination where these smaller packets are put back together into the 15,000-octet file.

While these packets are in transit, however, the network does not know, or care, that there is any relation between them. It is possible that packet 24 will arrive before packet 11. It is also possible that somewhere in the network an error will occur, and a packet will not be forwarded at all. In this case, the missing packet must be resent by the source.

Two separate protocols are involved:

• TCP is responsible for breaking up the message into packets, reassembling them at the other end, resending anything that gets lost, and putting these packets back in the right order.

• IP is responsible for routing individual packets from source to destination across the internetwork.

The interface between TCP and IP is simple: TCP hands IP a packet with a destination network address. IP does not know, nor care, how this packet relates to any packet before it or after it.

TCP/IP is based on the presumption that there are a large number of independent networks connected together by routers, switches, and gateways. The end-user should be able to access hosts or other resources on any of these networks with their packets often passing through a dozen different networks before reaching their final destination. The routing of these packets should be invisible to the end-user because as far as the user is concerned, all that is needed to access another system is an Internet address, something that looks like 128.6.4.194. Internet addresses are a 32-bit number, often written as four decimal numbers, each representing 8 bits of the address.

NOTE

The term octet is used by Internet documentation for 8-bit parts. The term byte is not used because TCP/IP is supported by some computers that have byte sizes other than 8 bits.

The structure of the Internet (network) address provides some information about how to get to the host. For example, 128.6.4.194 is broken down as follows:

• 128.6 is the network number assigned by a central authority (IANA) to Rutgers University.

• Rutgers uses the next octet to indicate which of the campus Ethernets is involved. 128.6.4 is the address of an Ethernet network used by the Computer Science Department.

• The last octet allows for up to 254 systems on each Ethernet network.

Internet users often refer to network systems by name rather than by Internet address. When a name is specified, the network software on the host, such as web browser, looks up the name in a database and is given the corresponding Internet address; this database is part of the Domain Name System (DNS).