inglese |
|
|||||
How an Ethernet Works
. Ethernet is not a single product, but rather a technical standard developed for network communications by Xeros, Dec and Intel, which the rest of the computer community have adopted.
. All nodes, clients and servers on an Ethernet network are attached to the LAN as branches off a common line in a bus conuration. Each node has a unique address. When a node (a PC, file server, a print server) needs to send data to another node, it sends the data, or message, through a network card installed in an expansion slot.
. The card listens to be sure that no other signals are being transmitted along the network card's transceiver. Each node's network connection has its own transceiver.
. The transceiver broadcasts the message in both directions so that it reaches all other nodes on the network. The message includes the addresses of the message destination and source, packets of data to be used for error checking, and the data itself.
. Each node along the bus network inspects the addressing information contained in the message. Nodes to which the message is not addressed ignore it.
. If a node detects its own address in a message, instead, it reads the data, checks for errors, and sends an acknowledgement to the sender, using the sender's address, which was included as part of the incoming message.
. When two nodes send messages simultaneously, the collision of the two messages creates a recognizable electrical interference pattern that travels along the bus. The request results in a "busy line" signal, which is detected by the senders.
. The first sender to detect a collision sends a special signal that jams the network so that all nodes will know that the network is blocked. Transmission from all nodes is halted, and each node waits a random length of time before trying to resend its message. The process repeats itself until one of the nodes sends its message without encountering another node message.
. Switches and routers use the information to determine where to forward the packet. In a network joined by hubs themselves check the address to determine which packets to pay attention to and which to ignore.
¥ Costly resources such as printers can be shred by all of the computers. This means that better quality printing is available to everyone because one or more expensive, lower specification models.
¥ Central backing store can be provided in one place (the dedicated file server), so all work is saved together. An individual user can load his or her work on any computer on the network and data stored on the server can be accessed from any terminal.
¥ Central backup can take place automatically at regular intervals. A user will be able to retrieve work that has been deleted by mistake.
¥ Data can be shared across the network. For example, this would allow several people to work on the same project.
¥ Software can be shared and upgrading would be easier. This gives continuity to the workplace.
¥ If the data being shared is in a database, several people will be able to use the database at the same time, but they will not be able to edit the same record at the same time. When a record is opened by one user, it is locked so that the other people cannot try to edit it at the same time. This avoids the confusion that would result if several people were trying to edit data simultaneously. Once the first user has completed and saved the operation, the record is unlocked again.
¥ Local e-mail messages can be sent to people working at other terminals on the network. This can save time and ensures that messages get to the right place.
¥ There may be an Intranet. This works like a World Wide Web, with es of information. The difference is that the es can be accessed only over the LAN. As it does not involve phone links, an Intranet is free.
¥ The use of e-mail where the network can lead to problems of time waiting as people send messages that do not relate to work. Some companies discourage the use of in house e-mail because they believe that it is better for employees to communicate face to face.
¥ Where a lot of terminals are served by only one or two printers, long print queues may develop, causing people to have to wait for printed output.
¥ Network security can be a problem. If a virus gets into one computer, it is likely to spread quickly across the network because it will get into the central backing store.
¥ Users of the network have to use user names and passwords. Some users are not very good at keeping passwords secret, or they may use passwords that are easy to guess. As a consequence, other people can then log on to the network.
¥ Cabling can be expensive to buy and install. In a busy situation, cabling must often be placed under the floor so that people will not trip over it. If connecting cables are damaged, some sections of the network can become isolated. They will not be able to communicate with the rest of the network.
The advantages and disadvantages of using WANs are very similar to those for LANs, but on a larger scale. Security is even more important particularly where WANs are connected to the public telephone system.
Privacy
|
© ePerTutti.com : tutti i diritti riservati
:::::
Condizioni Generali - Invia - Contatta