 |
|
 |
|
Technet FAQ - Dedicated
Internet Service Return to FAQ Table of Contents
How Can My Business Get Connected? If your business or office has a local area network or LAN, you can make good use of a "dedicated" Internet connection. This requires a data circuit, and special equipment, but allows your company to have an Internet address 24 hours a day, 7 days a week. You need a router and a CSU/DSU at your company's site. This equipment is connected to your LAN and to the data circuit on your company side. A router port and a CSU/DSU are also connected to the data circuit at an ISP. Once your company has a dedicated Internet connection, all employees using computers connected to your company's LAN can have e-mail access and web access from their desktops. Your company can even host its own web page on a computer connected to the LAN. Bandwidth is potential capacity, with a limit. Consider a drinking straw, a garden hose, a fire hose, and a city water line. As you proceed through those transmission devices, you get greater capacity and larger pumping mechanisms. So, more water can flow, per second, through a water main than it can through a drinking straw. But, as more people use the resource, there is the possibility that service to all users will begin to degrade. So it is with the wires and cables that carry Internet traffic. A 1.54 Megabit T1 line has more bandwidth than a 28, 800 bit per second dial-up line and so can handle more users simultaneously But it can slow down if many people use it at once, especially if their use is graphically oriented. Dedicated bandwidth is used much more efficiently than dial-in bandwidth because the protocols used are different. A 56 Kbps circuit will usually support 40 to 50 simultaneous connections that are performing primarily text-based activities such as reading e-mail, downloading text files, or telneting to another site. This same circuit will only support 15 to 20 simultaneous connections that are performing more graphics oriented activities like accessing the World Wide Web, downloading graphics files, or running other graphics applications across the Internet. Most of the time, there is a mix of text and graphics activities being performed. A T1 circuit is approximately 24 times the bandwidth of a 56 Kbps circuit, but because of the higher efficiency utilized may actually support 30 to 40 times as many connections. Fractional T1 circuits in various sizes may also be available from your ISP. These may come in 1/12 T1, 1/6 T1, 1/3 T1 or other fractions of the total T1 bandwidth. A router is a special kind of computer that is connected by two or more links to other parts of the Internet. It maintains tables of Internet addresses, and for each a piece of data (known as a "packet") it receives, the router determines the best way for this packet to get to its destination and sends it along the appropriate link. There is a router at every location where any of the thousands of links that make up the Internet come together. The routers on the Internet backbones are very specialized, expensive units from vendors that can forward millions of packets a second. Frame Relay is a special kind of leased line that can used for full-time Internet connections. Ordinary leased lines, called "Digital Data Service" or DDS, use equipment dedicated to a particular leased line, and these resources are wasted when the leased line is idle. DDS is often called "Dedicated Digital Service" because of the dedicated nature of the circuit. Frame Relay allows telephone company switch resources to be shared dynamically by several leased line subscribers, usually resulting in substantial savings when compared to DDS. Frame relay resources can also be shared at the end points of the circuit, often resulting in further savings on the routers and other equipment needed to use the leased line. However, if frame relay resources are extremely busy at a particular instant, some data frames may be discarded. Since TCP/IP automatically resends lost frames, this is not a serious problem at low rates of loss. Discard rates are prevented from growing too high by a guaranteed resource level called the "committed information rate" or CIR. Traffic at or below the CIR is always transmitted; only bursts that exceed the CIR are eligible to be discarded. The CIR is usually one half the total circuit capacity. Frame relay is not available at every telephone company central office in New Mexico, but traffic may be transmitted to a frame relay capable CO using DDS service. Anyone who wants to can become an ISP. Since most small communities do not have local dial-in lines from the larger Internet companies such as America On-line or Compuserve, local ISPs have found that this is a lucrative service they can provide. Larger communities can support multiple local ISPs, as well as national ones. Often a local ISP provides better service at lower cost, and provides a great deal more technical support. You will need to have a dedicated Internet connection with a router and CSU/DSU, a "host" computer or server, and some way to have your customers connect to your services, such as dial-in lines and modems. You can also provide dedicated connections to schools and businesses, web pages for your customers, and offer Internet training classes. The Internet is made up of separate, cooperating organizations which coordinate with each other on technical issues. The InterNIC is the organization that oversees this coordination, particularly in the areas of IP address allocation and domain name registration. The InterNIC insures that IP addresses and domain names on the Internet are unique, preventing conflicts that would result in interruption of service. The InterNIC is also one of the many places on the Internet where the technical standards that allow the Internet to function are published. In most cases, your Internet Service Provider will work with the InterNIC on your behalf as part of your Internet service. An intranet is simply a private or semi-private internal company network that uses standard Internet technologies and tools. Intranets can be very large and span wide geographic areas, but all the elements of the intranet are directly controlled by the organization using it. In contrast, the Internet is controlled by many organizations, each maintaining their own piece, with varying policies and practices. An example of an intranet application would be a web page for company employees with phone numbers, personnel policies, memos and other information of interest to employees, but not made available to the general public. Intranets can be private or isolated from the Internet, or semi-private, where a limited exchange of information is allowed with the world-wide Internet. A firewall is designed to allow you to closely monitor and control access to and from the Internet when you have a full-time Internet connection. Firewalls allow you to increase your security beyond the basic level that is built into most modern computer operating systems. Security-aware system administration practices such as enforcing a reasonable password policy, preventing access to sensitive information by inappropriate users, controlling physical access to computers, and routine monitoring of system logs often provide an adequate level of security. Highly critical or secret information may require a firewall for even more closely guarded access. It is important to note that a firewall does not replace the system administration practices already mentioned, but depends on them. It is also important to note that a firewall is largely ineffective in protecting critical information from inappropriate internal users. Firewalls can also cause delays, increased complexity, and blocked resources for internal users accessing the Internet. However, if basic security is insufficient to protect critical information on your computers, a firewall is a valuable additional investment. Each organization must decide what level of security is required, and if a firewall would be valuable in implementing it. This determination is based on the amount of critical or secret data kept online, and the level of resources that can be devoted to protecting it. The largest risk is often from internal users. Return to FAQ Table of Contents
|
|
|
