1. CS 105 “Tour of the Black Holes of Computing”
Internetworking
Topics
Client-server programming model
Networks
Internetworks
Global IP Internet
IP addresses
Domain names
Connections
net1.ppt

2. A Client-Server Transaction
Every network application is based on the client-server model:
A server process and one or more client processes
Server manages some resource.
Server provides service by manipulating resource for clients.
Clients and servers are processes running on hosts - can be same or different hosts
Different hosts require interconnection, I.e., a network
1. Client sends request
Client
process
Server
process
Resource
4. Client
handles
response
3. Server sends response
2. Server
handles
request

3. Computer Networks
A network is a hierarchical system of boxes and wires organized by geographical proximity
LAN (local area network) spans building or campus
Ethernet is most prominent example
(wireless) has become important
WAN (wide-area network) spans country or world
Typically high-speed point-to-point copper or fiber lines
Also microwave and satellite links in some situations
An internetwork (internet) is an interconnected set of networks
Global IP Internet (uppercase “I”) is most famous example of an internet (lowercase “i”)
Let’s look at how to build an internet from ground up

4. Suppose You Want to Build A Network
What available technologies would serve as building blocks?
What software architecture would you use?
How would you incorporate the future?
What principles would you enforce?
Who would manage it?
Who would control it?

5. Computer Networks Fundamental Requirement:
A Computer network must provide general, cost-effective, fair, robust, secure, and high performance connectivity among a large number of computers.
Internetworking
Abstraction that deals with complexity of multiple underlying communication technologies

6. Hardware Org of a Network Host
CPU chip
register file
ALU
system bus
memory bus
main
memory MI
I/O
bridge
Expansion slots
I/O bus
USB
controller
graphics
adapter
disk
controller
network
adapter
mouse
keyboard
monitor
disk
network

7. Original LAN A collection of hosts attached to a single common wire:
Ethernet now most known
Token Ring, X25 alternatives
...
host
host
host
Ethernet

8. Lowest Level: Ethernet Segment
Ethernet segment consists of a collection of hosts connected by wires (twisted pairs) to a hub/switch - replaces a common ‘wire’ or ‘bus’
Spans room or floor in a building.
Operation
Each Ethernet adapter has a unique 48-bit (MAC) address.
Hosts send bits to any other host in chunks called frames.
Hub slavishly copies each bit from each port to every other port.
Every adapter sees every bit; chooses which frames to hand to the system.
Hub replaces wire
Hub Alternative: switch copies bits only to proper destination
host
host
host
100 Mb/s
100 Mb/s
hub
Physical
Ports/connections

9. Next Level: Bridged Network Segment
Spans building or campus.
Bridges/switches cleverly learn which hosts are reachable from which ports and then selectively copy frames from port to port. How? Frames have source and destination addresses (board)…. A B
host
host
host
host
host X
hub
bridge
hub
100 Mb/s
100 Mb/s
1 Gb/s
host
host
100 Mb/s
100 Mb/s
hub
bridge
hub Y
host
host
host
host
host C

10. Conceptual View of LANs
For simplicity, hubs, bridges, and wires are often shown as a collection of hosts attached to a single wire:
...
host
host
host

11. Next Level: internets
Multiple incompatible LANs can be physically connected by specialized computers called routers (gateway).
The connected networks are called an internet.
...
...
host
host
host
host
host
host
LAN 1
LAN 2
router
router
router
WAN
WAN
LAN 1 and LAN 2 might be completely different technologies, totally incompatible LANs (e.g., Ethernet and ATM)

12. Notion of an internet Protocol
How is it possible to send bits across incompatible LANs and WANs?
Solution: protocol software running on each host and router that smoothes out the differences between the different networks.
Implements an internet protocol (i.e., set of rules) that governs how hosts and routers should cooperate when they transfer data from network to network.
TCP/IP is the protocol (family) for the global IP Internet.

13. What Does an internet Protocol Do?
1. Provides a naming scheme
An internet protocol defines a uniform format for host addresses.
Each network node (host, switch, router) is assigned at least one of these internet addresses that uniquely identifies it.
2. Provides a delivery mechanism
An internet protocol defines a standard transfer unit (datagram, frame, message, packet)
Packet consists of header and payload
Header: contains info such as packet size, source and destination addresses.
Payload: contains data bits sent from source host.
Encapsulation - key to network messages

14. Transferring Data via an internet
Host A
Host B
client
server
(1)
(8)
data
data
protocol
software
protocol
software
internet packet
(2)
(7)
data PH
data PH
LAN1
adapter
LAN2
adapter
Frame
Router
(3)
(6)
data PH
FH1
data PH
FH2
LAN1
adapter
LAN2
adapter
LAN1
LAN2
LAN2 frame
(4)
data PH
FH1
(5)
data PH
FH2
different
media
protocol
software

15. Other Issues We are glossing over a number of important questions:
What if different networks have different maximum frame sizes? (segmentation)
How do routers know where to forward frames?
How are routers informed when the network topology changes?
What if packets get lost?
These (and other) questions are addressed by the area of systems known as computer networking: CS 125.

16. Global IP Internet Most famous example of an internet.
Administered by a Unique Group - IETF
Based on the TCP/IP protocol family
IP (Internet protocol) :
Provides basic naming (addressing) scheme and unreliable, best effort delivery capability of packets (datagrams) from host-to-host.
UDP (Unreliable Datagram Protocol)
Uses IP to provide unreliable datagram delivery from process-to-process.
TCP (Transmission Control Protocol)
Uses IP to provide reliable byte streams from process-to-process over connections.
Accessed via a mix of Unix file I/O and functions from the sockets interface.

17. Hardware and Software Org of an Internet Application
Internet client host
Internet server host
Client
Server
User code
Sockets interface
(system calls)
TCP/IP
TCP/IP
Kernel code
Hardware interface
(interrupts)
Network
adapter
Hardware
and firmware
Network
adapter
Global IP Internet

18. Basic Internet Components
An Internet backbone is a collection of routers (nationwide or worldwide) connected by high-speed point-to-point networks.
A Network Access Point (NAP) is a router that connects multiple backbones (sometimes referred to as peers).
Regional networks are smaller backbones that cover smaller geographical areas (e.g., cities or states)
A point of presence (POP) is a machine that is connected to the Internet.
Internet Service Providers (ISPs) provide dial-up or direct access to POPs.

19. The Internet Circa 1993
In 1993, the Internet consisted of one backbone (NSFNET) that connected 13 sites via 45 Mbs T3 links.
Merit (Univ of Mich), NCSA (Illinois), Cornell Theory Center, Pittsburgh Supercomputing Center, San Diego Supercomputing Center, John von Neumann Center (Princeton), BARRNet (Palo Alto), MidNet (Lincoln, NE), WestNet (Salt Lake City), NorthwestNet (Seattle), SESQUINET (Rice), SURANET (Georgia Tech).
Connecting to the Internet involved connecting one of your routers to a router at a backbone site, or to a regional network that was already connected to the backbone.

20. NSFNET Internet Backbone
source:
US Centric, only one backbone

21. Enter Al Gore Myth: Al Gore claimed to have invented the Internet
Fact: In a 1999 interview, Al Gore said, “During my service in the United States Congress, I took the initiative in creating the Internet”
Fact: Dave Farber, Vint Cerf, and Bob Metcalfe have all supported the statement
Fact: Al Gore introduced and supported many bills funding the shift from a primarily US research network to a worldwide commercial one
Farber: “The guy used an inappropriate word. If he had said he was instrumental in the development of what it is now, he'd be accurate.”

22. Current NAP-Based Internet Arch
In the early 90’s commercial outfits were building their own high-speed backbones, connecting to NSFNET, and selling access to their POPs to companies, ISPs, and individuals.
In 1995, NSF decommissioned NSFNET, and fostered creation of a collection of NAPs to connect the commercial backbones.
Currently in the US there are about 50 commercial backbones connected by ~12 NAPs (peering points). NANOG
Similar architecture worldwide connects national networks to the Internet.

23. Abstracted Internet Hiearchy
Private
“peering”
agreements
between
two backbone
companies
often bypass
NAP
NAP
NAP
NAP
Colocation
sites
Backbone
Backbone
Backbone
Backbone
POP
POP
POP
POP
POP
POP
POP T3
Regional net
ISP
Big Business
POP
POP
POP
POP
POP
POP
POP
dialup
dialup T1 T1
ISP (for individuals)
Claremont Collegs
Pgh employee
DC employee

24. Network Access Points (NAPs)
Note: Peers in this context are
commercial backbones..droh
Source: Boardwatch.com

25. Source: Boardwatch.com
MCI/UUNET Global Backbone Example of Nested Internet
Source: Boardwatch.com

26. A Programmer’s View of the Internet
1. Hosts are mapped to a set of 32-bit IP addresses.
Class structure: A, B, C, Now Cider
Running out of IPv4 addresses, no more in general pool, allocated to locate registeries
2. The set of IP addresses is mapped to a set of identifiers called Internet domain names.
is mapped to
is mapped to
3. A process on one Internet host can communicate with a process on another Internet host over a connection -- IP Address, Port Number

27. 1. IP Addresses 32-bit IP addresses are stored in an IP address struct
IP addresses are always stored in memory in network byte order (big-endian byte order)
True in general for any integer transferred in a packet header from one machine to another.
E.g., the port number used to identify an Internet connection.
/* Internet address structure */
struct in_addr {
unsigned int s_addr; /* network byte order (big-endian) */ };
Handy network byte-order conversion functions (NoOps on some machines) – x86 is little-endian
htonl: convert long int from host to network byte order.
htons: convert short int from host to network byte order.
ntohl: convert long int from network to host byte order.
ntohs: convert short int from network to host byte order.

28. Dotted Decimal Notation
By convention, each byte in a 32-bit IP address is represented by its decimal value and separated by a period
IPv4 address 0x8002C2F2 =
IPv6 address 2001:1878:3010:9023:2186:8bff:fef9:a407
Functions for converting between binary IP addresses and dotted decimal strings:
inet_aton: converts a dotted decimal string to an IP address in network byte order.
inet_ntoa: converts an IP address in network byte order to its corresponding dotted decimal string.
“n” denotes network representation. “a” denotes application representation.

29. 2. Internet Domain Names unnamed root TLD - Top Level Domains
First-level domain names
mil
edu
gov
com
mit
hmc
berkeley
amazon
Second-level domain names cs
math
www
Third-level domain names
Wilkes
Turing

30. Properties of DNS Host Entries
Each host entry is an equivalence class of domain names and IP addresses.
Each host has a locally defined domain name localhost which always maps to the loopback address
Different kinds of mappings are possible:
Simple case: 1-1 mapping between domain name and IP addr:
turing.cs.hmc.edu. maps to
Multiple domain names mapped to the same IP address:
cs.hmc.edu and both map to
Multiple domain names mapped to multiple IP addresses:
aol.com and map to multiple IP addresses
Some valid domain names don’t map to any IP address:
for example: research.cs.hmc.edu

31. Domain Naming System (DNS)
The Internet maintains a mapping between IP addresses and domain names in a huge worldwide distributed database called DNS.
Conceptually, programmers can view the DNS database as a collection of millions of host entry structures, OLD STRUCT
Old Functions for retrieving host entries from DNS:
gethostbyname: query key is a DNS domain name.
gethostbyaddr: query key is an IP address.
/* DNS host entry structure */
struct hostent {
char *h_name; /* official domain name of host */
char **h_aliases; /* null-terminated array of domain names */
int h_addrtype; /* host address type (AF_INET) */
int h_length; /* length of an address, in bytes */
char **h_addr_list; /* null-terminated array of in_addr structs */ };

32. man gethostbyname struct hostent { The members of this structure are:
char *h_name; /* official name of host */
char **h_aliases; /* alias list */
int h_addrtype; /* host address type */
int h_length; /* length of address */
char **h_addr_list; /* list of addresses from name server */ };
#define h_addr h_addr_list[0] /* address, for backward compatibility */
The members of this structure are:
h_name Official name of the host.
h_aliases A NULL-terminated array of alternate names for the host.
h_addrtype The type of address being returned; usually AF_INET.
h_length The length, in bytes, of the address.
h_addr_list A NULL-terminated array of network addresses for the host.
Host addresses are returned in network byte order.
h_addr The first address in h_addr_list; this is for backward com-
patibility.

33. Domain Naming System (DNS)
Internet tracks mapping between IP addresses and domain names in huge worldwide distributed database called DNS.
Conceptually, programmers can view DNS database as collection of millions of address information structures NEW STRUCT
New Functions for retrieving host entries from DNS:
getaddrinfo: query key is DNS domain name
getnameinfo: query key is IP address (V4 or V6)
/* Address information structure (DNS only has + entries) */
struct addrinfo {
int ai_flags; /* Various options */
int ai_family; /* + AF_INET or AF_INET6 */
int ai_socktype; /* Preferred socket type */
int ai_protocol; /* Preferred protocol */
size_t ai_addrlen; /* Length of address */
struct sockaddr *ai_addr; /* + Encoded IP address */
char *ai_canonname; /* + Canonical host name */
struct addrinfo *ai_next; /* Link to next answer */ };

34. man getaddrinfo
This structure can be used to provide hints concerning the type of socket that the caller supports or wishes to use. The caller can supply the following structure elements in hints:
ai_family The protocol family that should be used. When ai_family
is set to PF_UNSPEC, it means the caller will accept any
protocol family supported by the operating system.
ai_socktype Denotes the type of socket that is wanted: SOCK_STREAM,
SOCK_DGRAM, or SOCK_RAW. When ai_socktype is zero the
caller will accept any socket type.
ai_protocol Indicates which transport protocol is desired,
IPPROTO_UDP or IPPROTO_TCP. If ai_protocol is zero the
caller will accept any protocol.
ai_flags The ai_flags field to which the hints parameter points
shall be set to zero or be the bitwise-inclusive OR of
one or more of the values AI_ADDRCONFIG, AI_ALL,
AI_CANONNAME, AI_NUMERICHOST, AI_NUMERICSERV, AI_PASSIVE,
and AI_V4MAPPED.
AI_ADDRCONFIG If the AI_ADDRCONFIG bit is set, IPv4
addresses shall be returned only if an ….

35. A Program That Queries DNS - NEW
int main(int argc, char **argv) { /* argv[1] is a domain name */
struct addrinfo hints, *host, *firsthost = NULL;
struct sockaddr_in *addr;
char buf[80];
memset(&hints, 0, sizeof hints);
hints.ai_flags = AI_CANONNAME;
hints.ai_family = AF_UNSPEC; /* Or AF_INET or AF_INET6 */
if (getaddrinfo(argv[1], NULL, &hints, &firsthost) != 0)
exit(1);
printf("official hostname: %s\n", firsthost->ai_canonname);
for (host = firsthost; host != NULL; host = host->ai_next) {
addr = (struct sockaddr_in *)host->ai_addr;
printf("address: %s\n", inet_ntop(addr->sin_family,
&addr->sin_addr, buf, sizeof buf)); }
exit(0);

36. Querying DNS from the Command Line
Domain Information Groper (dig) provides a scriptable command line interface to DNS.
linux> dig +short kittyhawk.cmcl.cs.cmu.edu
linux> dig +short -x
KITTYHAWK.CMCL.CS.CMU.EDU.
linux> dig +short aol.com
linux> dig +short -x
aol-v5.websys.aol.com.

37. 3. Internet Connections
Clients and servers communicate by sending streams of bytes over connections:
Point-to-point, full-duplex (2-way communication), and reliable.
A socket is an endpoint of a connection
Socket address is an:: IPaddress/port pair
A port is a 16-bit integer that identifies a process:
Ephemeral port: Assigned automatically on client when client makes a connection request
Well-known port: Associated with some service provided by a server (e.g., port 80 is associated with Web servers)
A connection is uniquely identified by the socket addresses of its endpoints (socket pair)
(cliaddr:cliport, servaddr:servport)

38. Putting it all Together: Anatomy of an Internet Connection
Client socket address
:51213
Server socket address
:80
Client
Server
(port 80)
Connection socket pair
( :51213, :80)
Client host address
Server host address

39. Next Time
How to use the sockets interface to establish Internet connections between clients and servers

40. A Program That Queries DNS - OLD
int main(int argc, char **argv) { /* argv[1] is a domain name
char **pp; * or dotted decimal IP addr */
struct in_addr addr;
struct hostent *hostp;
if (inet_aton(argv[1], &addr) != 0 // got address
hostp = Gethostbyaddr((const char *)&addr, sizeof(addr),
AF_INET);
else
hostp = Gethostbyname(argv[1]); // got name
printf("official hostname: %s\n", hostp->h_name);
for (pp = hostp->h_aliases; *pp != NULL; pp++)
printf("alias: %s\n", *pp);
for (pp = hostp->h_addr_list; *pp != NULL; pp++) {
addr.s_addr = *((unsigned int *)*pp);
printf("address: %s\n", inet_ntoa(addr)); }