/ THE PHYSICAL AND LINK LAYERS 3 in Software Draw barcode 3/9 in Software / THE PHYSICAL AND LINK LAYERS 3

How to generate, print barcode using .NET, Java sdk library control with example project source code free download:

CHAPTER 5 / THE PHYSICAL AND LINK LAYERS 3 using barcode implement for software control to generate, create barcode 39 image in software applications. rfid Ethernet header barcode code39 for None Ehternet destination address Ethernet source address Frame Hardware Protocol Hw Prot type type type size size 2 2 1 1 Op 2. ARP packet Sende r Ethernet address 6 28 bytes Sender IP addr 4 Target Target Ethernet address IP addr 6 4. Figure 5.8 ARP packet. destination s Et hernet address. Thus we need to nd which Ethernet address corresponds to which IP address. This is done by ARP.

To determine the Ether address for an IP address, the source machine broadcasts an ARP request packet (Figure 5.8) using Ethernet address ff:ff:ff:ff:ff:ff (see p. 25).

All machines on the (local) network read the packet and the target machine recognizes that the packet is a request for its address. It responds with an ARP reply containing its hardware address (the other machines do nothing). The source machine reads this reply and extracts the Ethernet address.

The source machine can now use that Ethernet address to construct and send data packets to the intended target. This packet is embedded in an Ethernet frame, of course. The frame type for ARP packets is 0806.

The packet is, in more detail: Hardware type. This is 1 for an Ethernet address. Protocol type.

This is 0800 for an IP address. Hardware size, protocol size. The number of bytes each of these kinds of addresses occupies.

This is 6 for Ethernet and 4 for IP. Op. This is 1 for an ARP request and 2 for an ARP reply.

Address elds. The hardware and protocol addresses for the source and target. In an ARP request the target hardware address is not lled in; after all, that is what we are trying to nd out.

In an ARP reply, the target is the original sender and the sender hardware address eld contains the information we are interested in. Of course it would be stupid and a waste of network bandwidth to have an ARP for every packet sent, so instead a cache of mappings from IP to Ethernet addresses is kept by each host. Entries in this cache time out and are removed after 20 minutes, typically.

The next packet will require a fresh ARP. If no machine on the network has the requested address, or that machine is down, no reply will be forthcoming. If this is so and an ARP reply is not received after a couple of seconds, an error message will be sent to the application trying to make the IP connection.

The user might see a message like no such host , or host unreachable . Use the Unix command. arp -a to look at the cache. 5.4 / ARP Sometimes it is useful to broadcast an ARP reply even when nobody has asked for it. For example, when a new machine has joined the network or an existing machine has changed its IP address for some reason (see p. 114).

This is called a gratuitous ARP. All machines on the network are free to read any ARP reply and update their caches to re ect the given address association. This is most important in order to break old associations that are no longer valid but where the old association is still cached.

Otherwise hosts might try to send packets to the wrong hardware address. The ARP can be used in situations other than Ethernet and IP as it has parameterized elds and so can be used to associate pairs of any types of addresses, but it is by far most associated with Ethernet..

Bridging IEEE 802.1d There is a cleve Software bar code 39 r trick with ARP that can be used to extend networks: for example, making an Ethernet span a distance larger than the speci cations allow, or joining a wireless network to a wired network so they appear to be a single network. To do this we need a bridge. This is just a machine with two network interfaces, one on each network (Figure 5.

9). If host h1 wishes to send to host h2 it must determine its hardware address. So it does an ARP broadcast for h2.

The bridge sees this request and answers on behalf of h2 (it does a proxy ARP), but it answers with its own address b1. Now h1 sends the data to what it believes is the address of h2, but is actually b1. The bridge reads the packet, recognizes that it is destined for h2 and forwards it to the other network where it is safely received by h2.

Note that the forwarded packet will have the source hardware address b2 and destination hardware address h2. If h2 replies it either (a) uses the hardware address it got from the original packet, namely b2, or (b) does an ARP, when the bridge again proxies for h1 with its own address b2. In either case, the data packet is delivered to the bridge, which forwards it to h1.

All this is very transparent to h1 and h2 who think that they are on the same network. Thus this is sometimes called transparent bridging. Notice that if h1 is communicating with both h2 and h3 the ARP cache on h1 will show that h2 and h3 apparently have the same Ethernet address, namely b1.

This is not a problem. This works very well for joining a pair of networks, but is less suitable for larger collections of networks, in particular when there are multiple routes between hosts. In that case a more sophisticated method as given in the IEEE 802.

1d Ethernet Bridging standard can be employed..
Copyright © . All rights reserved.