1. © MMII JW RyderCS 428 Computer Networks1 Mapping Internet to Physical Addresses  2 machines on a physical network can only communicate if they know each other’s physical address (PA)  How does a router or host map an IP addr. to a PA?  2 machines with IP addresses I A and I B and physical addresses P A and P B  Devise a scheme so that high level programs can only work with IP address

2. © MMII JW RyderCS 428 Computer Networks2 Address Mapping  Must be done all along path from source to destination  2 cases  Last step of delivering a packet  Delivery to host on physical network  All other steps  Delivery to router on physical network  Problem known as the ‘address resolution problem’

3. © MMII JW RyderCS 428 Computer Networks3 Physical Addresses  2 types  Ethernet type - large fixed address  ProNet type - token ring - small easily configurable  Resolution difficult for Ethernet, easy for ProNet  ProNet uses small intergers, allows user to choose the PA when installing the board

4. © MMII JW RyderCS 428 Computer Networks4 Physical Addresses  Key - choose IP and PA that have some part of them overlapping  User has choice when configuring card!  Example: 192.5.48.3 and PA 3  Computing PA from IP becomes easy  P A = f (I A )

5. © MMII JW RyderCS 428 Computer Networks5 Dynamic Binding - ARP  No hope of encoding 48 bit addr into 32 bit IP addr  Use the Ethernet broadcast ability to solve the problem  No central DB and new hosts can be dynamically added  Host A wants to resolve IP addr I B

6. © MMII JW RyderCS 428 Computer Networks6 ARP  Host A broadcasts a special packet that asks the host with IP addr I B to respond with its PA P B  All hosts receive the packet. Only B recognizes its IP addr  Sends a reply with its PA  Host A uses the received PA to send the packet to host B

7. © MMII JW RyderCS 428 Computer Networks7 ARP  See figure 5.1 on page 80  To reduce communication costs hosts keep ‘ARP caches’ to maintain recently acquired IP to PA binding information  Cache information can become ‘stale’  Assume hosts A and B. Over time B crashes or leaves  No indication that host B not there  Use timeouts, typically 20 mins.

8. © MMII JW RyderCS 428 Computer Networks8 IP Addressing  All hosts on a given physical network share same prefix  network + host  Splitting IP addrs keep routing tables reasonable sizes  Class A - 8 network + 24 host  Class B - 16 network + 16 host  Class C - 24 network + 8 hosts

9. © MMII JW RyderCS 428 Computer Networks9 Addressing  Important - Individual sites have the right to modify addresses and routes within their intranet as long as it remains invisible to other sites

10. © MMII JW RyderCS 428 Computer Networks10 Addressing  Very important to have multiple physical networks use the same IP network address  To minimize the use of class B addresses we need to use as many class C addresses as possible

11. © MMII JW RyderCS 428 Computer Networks11 Transparent Routers  A network using a class A addr can be extended through a simple trick  Arrange a physical network to multiplex several host connections through a single host port  See figure 10.1 on page 149  LAN does NOT have its own IP prefix

12. © MMII JW RyderCS 428 Computer Networks12 Transparent Routers  Hosts attached to it are assigned addresses as if they are directly connected to the WAN  Transparent router de-multiplexes datagrams that arrive from the WAN assigning them to appropriate hosts  Uses a table of addresses

13. © MMII JW RyderCS 428 Computer Networks13 Transparent Routers  Divide IP addresses into parts and encode information into unused parts  ARPANET is 10.0.0.0 ==> 10.p.u.i  Network (10)  Port on destination (p)  Destination (i)  (u) is UNUSED !!!

14. © MMII JW RyderCS 428 Computer Networks14 Transparent Routers  Transparent router can assign one host 10.1.1.5 and 10.1.2.5  “Same” IP addr for 2 hosts on 2 separate networks  Advantage - Require fewer network addresses because LANs can share IP prefix  Disadvantage - Only works with class A on ARPANET

15. © MMII JW RyderCS 428 Computer Networks15 Proxy ARP  aka promiscuous ARP, the ARP hack  Maps single IP prefix into 2 physical networks  See figure 10.2 on page 150  Applies only to networks that use ARP to bind internet addresses to PAs

16. © MMII JW RyderCS 428 Computer Networks16 Proxy ARP  With 2 networks A and B and 1 router R  R answers ARP requests on each network for hosts on the other network  It gives its PA as the addr matching PB then routes datagrams correctly  “In essence, R lies about IP to physical address binding”

17. © MMII JW RyderCS 428 Computer Networks17 Subnet Addresses  Subnetting most common of 3 techniques  Subnetting is a required part of IP addressing  Main router is interface to WAN  Routes datagrams to specific internal physical networks  See figure 10.3 on page 152

18. © MMII JW RyderCS 428 Computer Networks18 Subnets  Example class B addr 128.10.0.0  Break internal network into several ‘independent’ class C networks  128.10.1.0 - 128.10.2.0 - 128.10.n.0  Gateway to WAN upon receiving datagram discerns which local network gets packet