This article is intended as an introduction to the basics of SME and home computing networks.
- History of small networks.
Originally, small networks were all that existed. The computers were huge, their capacity low and there were no standards that covered communication between them. Slowly, driven by the need to exchange information between different makes of computer, with different operating systems and in distant locations, proprietary communication standards and networks began to be developed and deployed. Mostly, they were created by computer suppliers, government and educational groups to allow them to work together. Then came a welter of networking and communication standards and the hardware and software to support them. Eventually, the dust settled and a series of de-facto standards emerged, based on Ethernet and TCP-IP, and with these came the Internet. - Why networks?
A network, in the computer sense, is a mechanism to allow multiple computers to interact with one another in a manner that each can understand. Nowadays, not all of these ‘computers’ are like the one on your desk: the General Purpose Personal Computer. Some of those non-GPPC devices you might have in your house could include smart meters for gas and electricity, telephone systems, burglar or fire alarms, games consoles, mobile telephones, TV and video devices: the list is long and growing daily. Originally, such devices may have used wireless or mains or dedicated cabling to communicate both within the home and beyond, but nowadays most are beginning to standardise on the Ethernet/TCP-IP model - although this model now includes both wireless and mains cable communication within its standards. - The nuts and bolts of designing a network.
The key questions that you need to answer before you can decide exactly what kind of network you need are:
- What do you want it to do?
This is the basic question and can be rather difficult to answer comprehensively. There are a number of things that most people want to:
- Allow your PC(s) to access Internet services;
- Allow your PC/telephone to access VoIP services (such as Skype) as well as the normal (such as BT or AT&T) 'pots' network;
- Allow non-PC systems (for example smart meters, games consoles, burglar alarms) to access their networked services.
- Allow your laptop to access your home computer remotely across the Internet;
- Allow other users to access services that you provide from your private network;
- Make secure private connections between multiple sites using (e.g.) Virtual Private Network connections;
- Where do you want to do it?
Most of the issues addressed in this series of articles relate to Local Area Networks (LANs). This is the network that covers a single building or group of buildings on a single site. Generally, the site will be small with direct cable access available across the whole site. It is possible to use a single LAN across multiple sites that are close to one another (an extended LAN) even if a physical cable interconnection is not practicable but this is not advisable except in special circumstances.
Where a site is very large (such as an airfield or a farm with widely-spaced buildings) or there are multiple, widely separated sites, then a Wide Area Network (WAN) should be used.
What are the differences? Basically, within a LAN, there may be little or no filtering of signals between computers. With a WAN, only specific types of data are transmitted across the WAN, meaning that a slower (and cheaper) connection medium is practicable.
Location affects another key aspect of most networks: how they connect to external services such as the Internet. In towns or high-density rural areas, most networks connect to the Internet, for instance, using telephone lines (ADSL - Asymmetric Digital Subscriber Line) or cable services. Outside these areas, if ADSL or cable are not available or practicable, then satellite links may be used. Another, rarer, alternative is point-to-point wireless links (normally microwave).
- How do you want to do it?
As with most technologies, there are a number of alternatives when implementing a solution:
- Simple and relatively cheap: using the minimum hardware and with no resilience to failure
- Moderately complex: with key systems duplicated to provide a moderate level of resilience to failure
- All singing, where every single point of failure is duplicated to provide maximum availability. This is also the maximum cost solution.
- What are your priorities?
You need to assess your own priorities. As in the old joke "I want it good, I want it cheap and I want it now." "OK, pick any two and get back to me!", you have to balance capability, service level, delivery and support times with cost to get the best system you feel you can afford. It is possible to spend almost infinite amounts of money to get an all-singing system delivered in a week, but is it worth it to you?
Remember, costs fall into two main categories:
- Immediate, up front costs, for hardware, software and installation/commissioning;
- Ongoing costs for hardware and software upgrades, support and maintenance
It is poor financial planning to spend a huge amount on up-front costs if you haven’t budgeted for ongoing costs. Similarly, it is a bad deal if you scrimp on up-front costs and then have to pay out a fortune for replacements, repairs and other support. Worst of all is spending lots up front and still having to pay out a fortune for support...
- What do you want it to do?
No comments:
Post a Comment