If you are working on setting up a network, here are a few tips for you.
When building a network, it can be either two or more computers, or PDAs sharing data by way of a transport medium. This configuration can include a small business network in one room, two personal Digital Assistance (PDA’s) communicating by way of infrared, or a world-wide network connecting a million users, such as the internet. Networking on the internet has its own name called Internetworking.
Networks allow you to share data quickly, not like in years gone by where users had to place files on to a floppy disk, or print them, then deliver or e-mail them to their destination. All networks need a standard, or protocol – rules that let data flow properly from one network interface card to another. Each network device must have an NIC, often called a network adapter card, which is the interface between the computer and the network.
The NIC generally resides in the “motherboard” expansion slot and communicates the computer through an NIC device driver. Networking cable connects the NIC to the network, and vary for Ethernet and token-ring networks. NIC operate at the data-link layer of the OSI/RM. An NIC can be attached to a computer by any of the following:
- Paripheral component interconnect (PCI) card.
- Universal Serial Bus.
- IEEE 1394 (FireWire) device.
- Industry Standard Architecture (ISA) card.
The structure by which communications is effected, is the Open Systems interconnection (OSI) reference model whereby you send information as the client to the Application Layer, which in-turn passes it through the Presentation, Session, Transport, Network, Data-Link and Physical layers. It is then passed along a copper wire to the servers OSI/RM, arriving at the servers Application layer where it is processed. All information sent across the network is done in packets, containing a header, real data and a trailer.
The protocol used to allow computers to communicate is the Transmission Control Protocol/ Internet protocol (TCP/IP) which is default on most systems, including the current windows family operating systems, Novell Netware and all flavours of Unix. Although it is configured on private networks, TCP/IP is needed for internet access. Other protocols include; UDP, IP, and IPX/SPX.
Two geographical areas exist in a network; The Local area network (LAN), confined to a geographical area, and a Wide Area Network (WAN), that is more expansive, such as a state or a country, allowing users to share files and services. The primary difference between a LAN and a WAN, is that a WAN involves two separate networks, using either a router or a switch. The Internet is an extremely large WAN.
One of the oldest existing networks – the public switched telephone network (PSTN), connecting millions of users, and remains the cornerstone of internet-working. Since the inception of the telephone, voice has been carried over circuit-switched connections of the PSTN. Originally all telephone service were analog. Today, however the network in entirely digital except for the part that extends from the central office of the telephone company to the user.
To exchange data over a public telephone network, you can use a dial-up connection, a modem device that translates, or modulates, a digital signal coming from your computer into an analog signal that can be carried over a telephone line. The modem is attached to the receiving computer, demodulating the analog signal back in to a digital one. PSTN, a hybrid network, is an part of the internet infrastructure, because it furnishes most of the long distance connections. Most internet service providers (ISP’s) pay long distance providers for access to telephone lines.
Originally, networks were operated on a centralized, or mainframe model which usually limited networks to large, well-funded institutions, such as universities and Fortune 500 companies. By the late 1980’s however, many business networks adopt the client/server model using a more modular approach, allowing small to medium business to create powerful network solutions.
The first practical network solution was the mainframe (or Centralized Computing). This approach used central servers, or mainframes, and remote terminals. They were usually diskless, or “dumb” stations that could only request information. Most information processing occurred on the “back-end” (the server) not the “front end” (the client).
The client/server model, also called distributing computing is the way forward. Also called distributing computing, it attempts to cut network slowdown by dividing processing tasks between client (the front end) and the server (the back-end). The back end computer is generally more powerful than the front end and responsible for storing and presenting information.
The advantages of client/server in addition to shared task processing, client/server benefits include a modular approach to computing, because the client/server model allows you add new system components. You ar not limited to one solution. With the advent of open standards such as TCP/IP and ODBC heterogeneous systems can work together more effectively.
All networks consist of the same three basic elements, Protocols (communication rules on which all network elements must agree. Transmission Media (media that makes all networking elements to interconnect) and Network Services (resources – (printers) that are shared with all network users. Two basics types of network exist; peer-to-peer and server-based. The Enterprise network combines peer-to-peer and server-based types. Microsoft Peer-to-Peer is a legacy model in which Microsoft-based systems communicate with one another without using a centralised system to control authentication and access. They tend to be less expensive and easier to work with than server-based networks. However they are less secure, support fewer users and experience more problems with file system management. Windows 95, 98, Millennium Edition and Windows XP all support peer-to-peer networking.
A Server-based network is a configuration of nodes that offer resources to other hosts on the network. These are called servers, and they include printers, applications and documents. These server-based networks offer user security because a central database can track the resources that users can access. However they can be expensive with a full-time administrator. Example of server nodes include; Server, File, Mail, Web and Database servers. Client/server networks include; Novell NetWare, and Unix/Linux, Microsoft LAN Manager and Microsoft Windows NT/2000 and 2003 servers. Authentication comes in two ways; Centralised, and decentralised.
The basic configurations used to wire networks and called Topologies. They include; bus, star, ring and hybrid.
Now we go on to Network Operating Systems, that manage resources on a network and offer services to one or more clients. An N.O.S manages multiple users on a network; provide file and print servers, web, FTP and e-mail, and network security. An N.O.S lets clients access remote drives (those not on the user’s computer). Part of the NOS runs from the client, and part from the server. The most popular NOS are; Microsoft Windows, Unix/Linux and Novell NetWare. A term called Interoperability makes it easier for corporations with different clients and servers to create a network, even though they have different operating systems.
The NT family of networks operating systems began in 1993, with the versions including;
- Windows NT 3.1
- Windows NT 3.5
- Windows NT 3.51
- Windows NT4.0
- Windows 2000
- Windows Server 2003