The Executive's Basic Guide to Intranets

4. How Does an Intranet Work with Existing Networks?

Generally speaking, intranets are constructed by implementing a common language --the Internet's TCP/IP protocol-- across the enterprise, to enable the exchange of information between all existing (as well as newly developed) networks and network devices.

At its core, the basic network infrastructure needed to support an intranet consists of networking devices --hubs, switches, and routers-- all of which understand the TCP/IP protocol that is being used to transmit data across divergent networks. Hubs provide connectivity for desktops, printers, and servers, and are used to build LANs. While LANs are connected via switches and/or routers. At the desktop level, Network Interface Cards (NICs) are added to each PC to enable it to talk to other equipment on the network. All of these devices work in tandem to move data throughout an organization as well as to the Internet or other private networks.

Many organizations have "legacy systems" that need to be incorporated into an intranet. Some of these legacy systems were developed before the advent of LANs and therefore may not be capable of interpreting or speaking TCP/IP. Most often these are older mainframe or computer systems running mission critical applications crunching sensitive company information. Environments using the Systems Network Architecture (SNA) --a suite of protocols used to connect IBM mainframe systems-- are a typical example of a proprietary (non-standard) protocol that requires a gateway in order to be accessed by users on a TCP/IP network. IPX (Internet Packet Exchange) is another example of a widely used proprietary protocol developed by Novell that requires a special technique called "tunneling" that enables NetWare networks to send and receive data to and from TCP/IP networks. Some older versions of NetWare do not support this technique and also require a gateway to enable data transmission.

1997 Copyright 3Com