• Aucun résultat trouvé

Reverse Address Resolution Protocol (RARP)

Dans le document Illustrated TCP/IP (Page 140-144)

T

his protocol is used when a network station knows its MAC address but does not know its IP address. When would this happen? Diskless workstations are a good example.

Notice that RARP uses the ARP packet format and does not involve IP; therefore, this packet cannot be routed. This protocol has been in use for some time, but there are other protocols that do a better job. This is one of the reasons that we use BOOTP and DHCP for address assignment because they can be forwarded over a router (with a

little assistance from the router). One problem with RARP is that like its cousin ARP, it does not use IP. Therefore, RARP is generally used only on a LAN.

The requesting client machine will send out a RARP request to a server located on the local segment that has the RARP server service running on it. This RARP server will respond to the request with that particular station’s IP address. Although the RARP server does not need to be located on the same cable segment or extended LAN, it is preferred. Some router vendors have enabled their routers to forward these requests and responses to other networks.

Reverse Address Resolution Protocol (RARP)

The packet format for a RARP packet is the same as for ARP. The only difference is that the field that will be filled in will be the sender’s physical address. The IP address fields will be empty. A RARP server will receive this packet, fill in the IP address fields, and reply to the sender—the opposite of the ARP process.

Other protocol similar to this are BOOTP and Dynamic Host Configuration Protocol (DHCP). DHCP is more powerful than RARP, but it does supply one of the same functions as RARP: resolving an IP address. Besides being less functional than DHCP, RARP only works on single subnets. RARP works at the datalink layer and therefore cannot span

subnets gracefully. DHCP can span subnets.

Previous Table of Contents Next

Illustrated TCP/IP by Matthew G. Naugle

Wiley Computer Publishing, John Wiley & Sons, Inc.

ISBN: 0471196568 Pub Date: 11/01/98

Previous Table of Contents Next

Chapter 66 Proxy ARP

P

roxy ARP protocol is not used much anymore, but it is still worth mentioning. IP was pretty well established when ARP came along, and some TCP/IP implementations did not support ARP. However, TCP/IP over LANs with subnets was being implemented and an interim solution was needed. This was the purpose of Proxy ARP (also known as ARP Hack). Proxy ARP is the ability of a router to be able to respond to an endstation (host) ARP request for a host that thinks the destination IP address is on the local LAN.

Therefore, if a host does not support subnet addressing, it could incorrectly mistake an IP subnet number for a host number. The router tricks the transmitting station into believing that the source station is on the local LAN.

Endstation A thinks host B is on the local LAN. Host B supports subnet addressing and endstation A does not. Deciphering the IP address, the first two fields (containing the network ID) are the same. Therefore, endstation A will send out a local ARP request packet when it should be submitting the packet to the router so that it can deliver the packet to the endstation. If the router has proxy ARP enabled, the router will answer for host B. The router, which supports subnetting, will look up the ARP request and then notice that the subnetwork address is in its routing table. The router responds for endstation B. Endstation A will receive this response and think it is from host B—there is nothing in the physical address of a packet to indicate where it came from. The host will then submit all packets to the router and the router will deliver them to

endstation A. This communication will continue until one end terminates the session.

Proxy ARP

Proxy ARP is a very useful protocol for those networks that have been using bridges to implement their IP network and are moving to a router environment. There are other situation for which proxy ARP is appropriate, but its use is waning. Today, most hosts on

a TCP/IP internet support subnet masking and most IP networks are using routers.

A potential problem in using proxy ARP is for those networks that implement the mechanism to ensure single IP addresses are on each network. Most TCP/IP

implementations allow users easy access to their network number (that is, they can change it with a text editor). This allows any hacker to change his or her number to another in order to receive datagrams destined for another host. Some implementations of TCP/IP will detect for this. Routers that implement proxy ARP will get caught, for they will answer for any station on a different network, thereby giving the impression that there is one physical address to multiple IP addresses. There is a trust on any IP network that IP addresses will not be arbitrarily assigned. There should be one IP address for each physical address on an internet.

Previous Table of Contents Next

Illustrated TCP/IP by Matthew G. Naugle

Wiley Computer Publishing, John Wiley & Sons, Inc.

ISBN: 0471196568 Pub Date: 11/01/98

Previous Table of Contents Next

Chapter 67

Dans le document Illustrated TCP/IP (Page 140-144)

Documents relatifs