1. Networking Alan L. Cox alc@cs.rice.edu T. S. Eugene Ng eugeneng@cs.rice.edu Some slides adapted from CMU 15.213 slides

2. Objectives Be exposed to the basic underpinnings of the Internet Be able to use network socket interfaces effectively Be exposed to the basic underpinnings of the World Wide Web Cox & NgNetworking2

3. 3 The 2004 A. M. Turing Award Goes to... Cox & NgNetworking

4. 4 The 2004 A. M. Turing Award Goes to... "For pioneering work on internetworking, including the design and implementation of the Internet's basic communications protocols, TCP/IP, and for inspired leadership in networking." The only Turing Award given to-date to recognize work in computer networking Bob KahnVint Cerf Cox & NgNetworking

5. 5 But at the Same Time... Source: Akamai Technologies, Inc. Daily, e.g. 110 attacks from China; 26 attacks from USA 2/24/2008 YouTube traffic mis-routed to Pakistan Q2/2008 1000s of Netherlands DSL customers lost service due to network configuration error 11/10/2008 CTBC (Brazil) black-holed all Internet traffic in some parts of Brazil 2/16/2009 Supro (Czech) routing messages triggered a Cisco router bug world-wide 5/2009 AfNOG (Africa) routing messages triggered buffer overflow in open-source routing software Quagga world-wide Cox & NgNetworking

6. 6 Telephony Interactive telecommunication between people Analog voice  Transmitter/receiver continuously in contact with electronic circuit  Electric current varies with acoustic pressure Over electrical circuits Analog/Continuous Signal Cox & NgNetworking

7. 7 Telephony Milestones 1876: Alexander Bell invented telephone 1878: Public switches installed at New Haven and San Francisco, public switched telephone network is born  People can talk without being on the same wire! Without SwitchWith Switch Cox & NgNetworking

8. 8 Telephony Milestones 1878: First telephone directory; White House line 1881: Insulated, balanced twisted pair as local loop 1885: AT&T formed 1892: First automatic commercial telephone switch 1903: 3 million telephones in U.S. 1915: First transcontinental telephone line 1927: First commercial transatlantic commercial service Cox & NgNetworking

9. 9 Telephony Milestones 1937: Multiplexing introduced for inter-city calls  One link carries multiple conversations Without Multiplexing With Multiplexing Cox & NgNetworking

10. 10 Data or Computer Networks Networks designed for computers to computers or devices  vs. communication between human beings Digital information  vs. analog voice Not a continuous stream of bits, rather, discrete “packets” with lots of silence in between  Dedicated circuit hugely inefficient  Packet switching invented Digital/Discrete Signal Cox & NgNetworking

11. 11 Major Internet Milestones 1960-1964 Basic concept of “packet switching” was independently developed by Baran (RAND), Kleinrock (MIT)  AT&T insisted that packet switching would never work! 1965 First time two computers talked to each other using packets (Roberts, MIT; Marill, SDC) MIT TX-2 SDC Q32 dial-up Cox & NgNetworking

12. 12 Major Internet Milestones 1968 BBN group proposed to use Honeywell 516 mini-computers for the Interface Message Processors (i.e. packet switches) 1969 The first ARPANET message transmitted between UCLA (Kleinrock) and SRI (Engelbart)  We sent an “L”, did you get the “L”? Yep!  We sent an “O”, did you get the “O”? Yep!  We sent a “G”, did you get the “G”? Crash! Cox & NgNetworking

13. 13 Major Internet Milestones 1970 First packet radio network ALOHANET (Abramson, U Hawaii) 1973 Ethernet invented (Metcalfe, Xerox PARC) Why is it called the “Inter-net”? 1974 “A protocol for Packet Network Interconnection” published by Cerf and Kahn –First internetworking protocol TCP –This paper was cited for their Turing Award 1977 First TCP operation over ARPANET, Packet Radio Net, and SATNET 1985 NSF commissions NSFNET backbone 1991 NSF opens Internet to commercial use Cox & NgNetworking

14. 14 Internet Hourglass Architecture Ethernet, WiFi, 3G, bluetooth,... IP TCP, UDP, … Email, Web, ssh,... Cox & NgNetworking

15. Cox & NgNetworking15 A Client-Server Transaction Most network applications are 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 Client process Server process 1. Client sends request 2. Server handles request 3. Server sends response 4. Client handles response Resource Note: clients and servers are processes running on hosts (can be the same or different hosts)

16. Cox & NgNetworking16 Network Hardware main memory I/O bridge Bus Interface ALU register file CPU chip system busmemory bus disk controller graphics adapter USB controller mousekeyboardmonitor disk I/O bus Expansion slots network adapter network

17. Cox & NgNetworking17 Computer Networks A network is a hierarchical system of boxes and wires organized by geographical proximity  Cluster network spans cluster or machine room Switched Ethernet, Infiniband, Myrinet, …  LAN (local area network) spans a building or campus Ethernet is most prominent example  WAN (wide-area network) spans very long distance A high-speed point-to-point link Leased line or SONET/SDH circuit, or MPLS/ATM circuit An internetwork (internet) is an interconnected set of networks  The Global IP Internet (uppercase “I”) is the most famous example of an internet (lowercase “i”)

18. Cox & NgNetworking18 Lowest Level: Ethernet Segment Ethernet segment consists of a collection of hosts connected by wires (twisted pairs) to a hub Operation  Each Ethernet adapter has a unique 48-bit 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 host sees every bit  Note: Hubs are largely obsolete Bridges (switches, routers) became cheap enough to replace them (don’t broadcast all traffic) host hub 100 Mb/s ports

19. Cox & NgNetworking19 Next Level: Bridged Ethernet Segment Spans room, building, or campus Bridges cleverly learn which hosts are reachable from which ports and then selectively copy frames from port to port host hub bridge 100 Mb/s host 1 Gb/s 1-10 Gb/s host bridge host bridge 1 Gb/s

20. Cox & NgNetworking20 Conceptual View of LANs For simplicity, hubs, bridges, and wires are often shown as a collection of hosts attached to a single wire: host...

21. Cox & NgNetworking21 Next Level: internets Multiple incompatible LANs can be physically connected by specialized computers called routers The connected networks are called an internet host LAN 1... host LAN 2... router WAN LAN 1 and LAN 2 might be completely different, totally incompatible LANs (e.g., Ethernet and WiFi, 802.11*, T1-links, DSL, …)

22. Cox & NgNetworking22 The Internet Circa 1986  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) In 1986, the Internet consisted of one backbone (NSFNET) that connected 13 sites via 45 Mbps T3 links 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

23. Cox & NgNetworking23 NSFNET Internet Backbone source: www.eef.org

24. Cox & NgNetworking24 After NSFNET Early 90s  Commercial enterprises began building their own high-speed backbones  Backbone would connect to NSFNET, sell access to companies, ISPs, and individuals 1995  NSFNET decommissioned  NSF fostered the creation of network access points (NAPs) to interconnect the commercial backbones

25. Cox & NgNetworking25 Current Internet Architecture

26. Cox & NgNetworking26 Level 3 Backbone

27. Cox & NgNetworking27 AT&T Backbone

28. Cox & NgNetworking28 Submarine Cabling

29. Cox & NgNetworking29 The 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 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 for the global IP Internet

30. Cox & NgNetworking30 What Does an internet Protocol Do? 1. Provides a naming scheme  An internet protocol defines a uniform format for host addresses  Each host (and 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 (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

31. Cox & NgNetworking31 Transferring Data Over an internet protocol software LAN1 adapter Host A data PHFH1 dataPH dataPHFH2 LAN1 LAN2 data PH FH1 dataPHFH2 (1) (2) (3) (4) (5) (6) (7) (8) internet packet LAN2 frame protocol software LAN1 adapter LAN2 adapter Router FH1 LAN1 frame dataPHFH2 protocol software LAN2 adapter Host B clientserver

32. Cox & NgNetworking32 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? We’ll leave the discussion of these question to computer networking classes (COMP 429)

33. Cox & NgNetworking33 Global IP Internet Based on the TCP/IP protocol family  IP (Internet protocol) : Provides basic naming scheme and unreliable 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

34. Cox & NgNetworking34 Organization of an Internet Application TCP/IP Client Network adapter Global IP Internet TCP/IP Server Network adapter Internet clientInternet server Sockets interface (system calls) Hardware interface (interrupts) User code Kernel code Hardware and firmware

35. Cox & NgNetworking35 A Programmer’s View of the Internet Hosts are mapped to a set of 32-bit IP addresses  128.42.1.125 (4 * 8 bits) A set of identifiers called Internet domain names are mapped to the set of IP addresses for convenience  www.cs.rice.edu is mapped to 128.42.1.125 A process on one Internet host can communicate with a process on another Internet host over a connection

36. Cox & NgNetworking36 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: 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

37. Cox & NgNetworking37 Dotted Decimal Notation By convention, each byte in a 32-bit IP address is represented by its decimal value and separated by a period IP address 0x8002C2F2 = 128.2.194.242 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 by order to its corresponding dotted decimal string  “ n ” denotes network representation, “ a ” denotes application representation

38. Cox & NgNetworking38 IP Address Structure IP (V4) Address space divided into classes: Special Addresses for routers and gateways (all 0/1’s) Loop-back address: 127.0.0.1 Unrouted (private) IP addresses:  10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16 Dynamic IP addresses (DHCP) Class A Class B Class C Class D Class E 0 1 2 3 8 16 24 31 0 Net IDHost ID 1 0 11 0 Net ID 11 0 1111 1 Multicast address Reserved for experiments

39. Cox & NgNetworking39 Internet Domain Names.net.edu.gov.com riceberkeleymit clearcs bell 128.42.151.14 unnamed root crystal 128.42.151.13 amazon www 72.21.210.11 First-level domain names Second-level domain names Third-level domain names

40. Cox & NgNetworking40 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: 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 */ };

41. Cox & NgNetworking41 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 127.0.0.1 Different kinds of mappings are possible:  Simple case: 1 domain name maps to one IP address: forest.owlnet.rice.edu maps to 10.130.195.71  Multiple domain names mapped to the same IP address: www.cs.rice.edu, ececs.rice.edu, and bianca.cs.rice.edu all map to 128.42.1.125  Multiple domain names mapped to multiple IP addresses: aol.com and www.aol.com map to multiple IP addresses  Some valid domain names don’t map to any IP address: for example: clear.rice.edu

42. Cox & NgNetworking42 A Program That Queries DNS 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) hostp = Gethostbyaddr((const char *)&addr, sizeof(addr), AF_INET); else hostp = Gethostbyname(argv[1]); 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)); }

43. Cox & NgNetworking43 Querying DNS Domain Information Groper (dig) provides a scriptable command line interface to DNS  Available on CLEAR  Lots of web interfaces (google “domain information groper”) unix> dig +short bianca.cs.rice.edu 128.42.1.125 unix> dig +short -x 128.42.1.125 bianca.cs.rice.edu. unix> dig +short google.com 72.14.207.99 64.233.187.99 64.233.167.99 unix> dig +short -x 64.233.167.99 py-in-f99.google.com.

44. Cox & NgNetworking44 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 IP address, 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)

45. Cox & NgNetworking45 Putting it all Together: Anatomy of an Internet Connection Connection socket pair (128.2.194.242:51213, 208.216.181.15:80) Server (port 80) Client Client socket address 128.2.194.242:51213 Server socket address 208.216.181.15:80 Client host address 128.2.194.242 Server host address 208.216.181.15

46. Cox & NgNetworking46 Clients Examples of client programs  Web browsers, ftp, telnet, ssh How does a client find the server?  The IP address in the server socket address identifies the host (more precisely, an adapter on the host)  The (well-known) port in the server socket address identifies the service, and thus implicitly identifies the server process that performs that service  Examples of well known ports Port 7: Echo server Port 23: Telnet server Port 25: Mail server Port 80: Web server

47. Cox & NgNetworking47 Using Ports to Identify Services Web server (port 80) Client host Server host 128.2.194.242 Echo server (port 7) Service request for 128.2.194.242:80 (i.e., the Web server) Web server (port 80) Echo server (port 7) Service request for 128.2.194.242:7 (i.e., the echo server) Kernel Client

48. Cox & NgNetworking48 Servers Servers are long-running processes (daemons)  Created at boot-time (typically) by the init process (process 1)  Run continuously until the machine is turned off Each server waits for requests to arrive on a well-known port associated with a particular service  Port 7: echo server  Port 23: telnet server  Port 25: mail server  Port 80: HTTP server A machine that runs a server process is also often referred to as a “server”

49. Cox & NgNetworking49 Server Examples Web server (port 80)  Resource: files/compute cycles (CGI programs)  Service: retrieves files and runs CGI programs on behalf of the client FTP server (20, 21)  Resource: files  Service: stores and retrieve files Telnet server (23)  Resource: terminal  Service: proxies a terminal on the server machine Mail server (25)  Resource: email “spool” file  Service: stores mail messages in spool file See /etc/services for a comprehensive list of the services available on a UNIX machine

50. Cox & NgNetworking50 Sockets Interface Created in the early 80’s as part of the original Berkeley distribution of Unix that contained an early version of the Internet protocols Provides a user-level interface to the network Underlying basis for all Internet applications Based on client/server programming model

51. Cox & NgNetworking51 Overview of the Sockets Interface Client Server socket bind listen accept rio_readlineb rio_writen close rio_readlineb connect rio_writen close Connection request EOF Await connection request from next client open_listenfd open_clientfd

52. Cox & NgNetworking52 Sockets What is a socket?  To the kernel, a socket is an endpoint of communication  To an application, a socket is a file descriptor that lets the application read/write from/to the network Remember: all Unix I/O devices, including networks, are modeled as files Clients and servers communicate with each other by reading from and writing to socket descriptors The main distinction between regular file I/O and socket I/O is how the application “opens” the socket descriptors

53. Cox & NgNetworking53 Socket Address Structures Generic socket address:  Address for connect, bind, and accept  C did not have generic (void *) pointers when the sockets interface was designed Internet-specific socket address:  Must cast (sockaddr_in *) to (sockaddr *) struct sockaddr { unsigned short sa_family; /* protocol family */ char sa_data[14]; /* address data. */ }; struct sockaddr_in { unsigned short sin_family; /* address family (always AF_INET) */ unsigned short sin_port; /* port num in network byte order */ struct in_addr sin_addr; /* IP addr in network byte order */ unsigned char sin_zero[8]; /* pad to sizeof(struct sockaddr) */ };

54. Cox & NgNetworking54 Echo Client Main Routine int main(int argc, char **argv) /* usage:./echoclient host port */ { int clientfd, port; char *host, buf[MAXLINE]; rio_t rio; host = argv[1]; port = atoi(argv[2]); clientfd = Open_clientfd(host, port); Rio_readinitb(&rio, clientfd); while (Fgets(buf, MAXLINE, stdin) != NULL) { Rio_writen(clientfd, buf, strlen(buf)); Rio_readlineb(&rio, buf, MAXLINE); Fputs(buf, stdout); } Close(clientfd); exit(0); }

55. Cox & NgNetworking55 Echo Client: open_clientfd int open_clientfd(char *hostname, int port) { int clientfd; struct hostent *hp; struct sockaddr_in serveraddr; if ((clientfd = socket(AF_INET, SOCK_STREAM, 0)) < 0) return -1; /* check errno for cause of error */ /* Fill in the server's IP address and port */ if ((hp = gethostbyname(hostname)) == NULL) return -2; /* check h_errno for cause of error */ bzero((char *) &serveraddr, sizeof(serveraddr)); serveraddr.sin_family = AF_INET; bcopy((char *)hp->h_addr, (char *)&serveraddr.sin_addr.s_addr, hp->h_length); serveraddr.sin_port = htons(port); /* Establish a connection with the server */ if (connect(clientfd, (SA *) &serveraddr, sizeof(serveraddr)) < 0) return -1; return clientfd; } This function opens a connection from the client to the server at hostname:port

56. Cox & NgNetworking56 open_clientfd (socket) socket creates a socket descriptor on the client  AF_INET : indicates that the socket is associated with Internet protocols  SOCK_STREAM : selects a reliable byte stream connection int clientfd; /* socket descriptor */ if ((clientfd = socket(AF_INET, SOCK_STREAM, 0)) < 0) return -1; /* check errno for cause of error */... (more)

57. Cox & NgNetworking57 open_clientfd (gethostbyname) The client then builds the server’s Internet address int clientfd; /* socket descriptor */ struct hostent *hp; /* DNS host entry */ struct sockaddr_in serveraddr; /* server’s IP address */... /* fill in the server's IP address and port */ if ((hp = gethostbyname(hostname)) == NULL) return -2; /* check h_errno for cause of error */ bzero((char *) &serveraddr, sizeof(serveraddr)); serveraddr.sin_family = AF_INET; bcopy((char *)hp->h_addr, (char *)&serveraddr.sin_addr.s_addr, hp->h_length); serveraddr.sin_port = htons(port);

58. Cox & NgNetworking58 open_clientfd (connect) Finally the client creates a connection with the server  Client process suspends (blocks) until the connection is created  After resuming, the client is ready to begin exchanging messages with the server via Unix I/O calls on descriptor clientfd int clientfd; /* socket descriptor */ struct sockaddr_in serveraddr; /* server address */ typedef struct sockaddr SA; /* generic sockaddr */... /* Establish a connection with the server */ if (connect(clientfd, (SA *)&serveraddr, sizeof(serveraddr)) < 0) return -1; return clientfd;

59. Cox & NgNetworking59 Echo Server: Main Routine int main(int argc, char **argv) { int listenfd, connfd, port, clientlen; struct sockaddr_in clientaddr; struct hostent *hp; char *haddrp; port = atoi(argv[1]); /* the server listens on the given port */ listenfd = open_listenfd(port); while (1) { clientlen = sizeof(clientaddr); connfd = Accept(listenfd, (SA *)&clientaddr, &clientlen); hp = Gethostbyaddr((const char *)&clientaddr.sin_addr.s_addr, sizeof(clientaddr.sin_addr.s_addr), AF_INET); haddrp = inet_ntoa(clientaddr.sin_addr); printf("server connected to %s (%s)\n", hp->h_name, haddrp); echo(connfd); Close(connfd); }

60. Cox & NgNetworking60 Echo Server: open_listenfd int open_listenfd(int port) { int listenfd, optval=1; struct sockaddr_in serveraddr; /* Create a socket descriptor */ if ((listenfd = socket(AF_INET, SOCK_STREAM, 0)) < 0) return -1; /* Eliminates "Address already in use" error from bind. */ if (setsockopt(listenfd, SOL_SOCKET, SO_REUSEADDR, (const void *)&optval, sizeof(int)) < 0) return -1;... (more)

61. Cox & NgNetworking61 Echo Server: open_listenfd (cont)... /* Listenfd will be an endpoint for all requests to port on any IP address for this host */ bzero((char *) &serveraddr, sizeof(serveraddr)); serveraddr.sin_family = AF_INET; serveraddr.sin_addr.s_addr = htonl(INADDR_ANY); serveraddr.sin_port = htons((unsigned short)port); if (bind(listenfd, (SA *)&serveraddr, sizeof(serveraddr)) < 0) return -1; /* Make it a listening socket ready to accept connection requests */ if (listen(listenfd, LISTENQ) < 0) return -1; return listenfd; }

62. Cox & NgNetworking62 open_listenfd (socket) socket creates a socket descriptor on the server  AF_INET: indicates that the socket is associated with Internet protocols  SOCK_STREAM: selects a reliable byte stream connection int listenfd; /* listening socket descriptor */ /* Create a socket descriptor */ if ((listenfd = socket(AF_INET, SOCK_STREAM, 0)) < 0) return -1;

63. Cox & NgNetworking63 open_listenfd (setsockopt) The socket can be given some attributes Handy trick that allows us to rerun the server immediately after we kill it  Otherwise we would have to wait about 30 secs  Eliminates “Address already in use” error from bind () Strongly suggest you do this for all your servers to simplify debugging... /* Eliminates "Address already in use" error from bind(). */ if (setsockopt(listenfd, SOL_SOCKET, SO_REUSEADDR, (const void *)&optval, sizeof(int)) < 0) return -1;

64. Cox & NgNetworking64 open_listenfd (init socket address) Next, we initialize the socket with the server’s Internet address (IP address and port) IP address and port stored in network (big-endian) byte order  htonl () converts longs from host byte order to network byte order  htons () convers shorts from host byte order to network byte order struct sockaddr_in serveraddr; /* server's socket addr */... /* listenfd will be an endpoint for all requests to port on any IP address for this host */ bzero((char *) &serveraddr, sizeof(serveraddr)); serveraddr.sin_family = AF_INET; serveraddr.sin_addr.s_addr = htonl(INADDR_ANY); serveraddr.sin_port = htons((unsigned short)port);

65. Cox & NgNetworking65 open_listenfd (bind) bind associates the socket with the socket address we just created int listenfd; /* listening socket */ struct sockaddr_in serveraddr; /* server’s socket addr */... /* listenfd will be an endpoint for all requests to port on any IP address for this host */ if (bind(listenfd, (SA *)&serveraddr, sizeof(serveraddr)) < 0) return -1;

66. Cox & NgNetworking66 open_listenfd (listen) listen indicates that this socket will accept connection ( connect ) requests from clients We’re finally ready to enter the main server loop that accepts and processes client connection requests int listenfd; /* listening socket */... /* Make it a listening socket ready to accept connection requests */ if (listen(listenfd, LISTENQ) < 0) return -1; return listenfd; }

67. Cox & NgNetworking67 Echo Server: Main Loop The server loops endlessly, waiting for connection requests, then reading input from the client, and echoing the input back to the client main() { /* create and configure the listening socket */ while(1) { /* Accept(): wait for a connection request */ /* echo(): read and echo input lines from client til EOF */ /* Close(): close the connection */ }

68. Cox & NgNetworking68 Echo Server: accept accept () blocks waiting for a connection request accept returns a connected descriptor ( connfd ) with the same properties as the listening descriptor ( listenfd )  Returns when the connection between client and server is created and ready for I/O transfers  All I/O with the client will be done via the connected socket accept also fills in client’s IP address int listenfd; /* listening descriptor */ int connfd; /* connected descriptor */ struct sockaddr_in clientaddr; int clientlen; clientlen = sizeof(clientaddr); connfd = Accept(listenfd, (SA *)&clientaddr, &clientlen);

69. Cox & NgNetworking69 Echo Server: accept Illustrated listenfd(3) Client 1. Server blocks in accept, waiting for connection request on listening descriptor listenfd clientfd Server listenfd(3) Client clientfd Server 2. Client makes connection request by calling and blocking in connect Connection request listenfd(3) Client clientfd Server 3. Server returns connfd from accept. Client returns from connect. Connection is now established between clientfd and connfd connfd(4)

70. Cox & NgNetworking70 Connected vs. Listening Descriptors Listening descriptor  End point for client connection requests  Created once and exists for lifetime of the server Connected descriptor  End point of the connection between client and server  A new descriptor is created each time the server accepts a connection request from a client  Exists only as long as it takes to service client Why the distinction?  Allows for concurrent servers that can communicate over many client connections simultaneously E.g., Each time we receive a new request, we fork a child to handle the request

71. Cox & NgNetworking71 Echo Server: Identifying the Client The server can determine the domain name and IP address of the client struct hostent *hp; /* pointer to DNS host entry */ char *haddrp; /* pointer to dotted decimal string */ hp = Gethostbyaddr((const char *)&clientaddr.sin_addr.s_addr, sizeof(clientaddr.sin_addr.s_addr), AF_INET); haddrp = inet_ntoa(clientaddr.sin_addr); printf("server connected to %s (%s)\n", hp->h_name, haddrp);

72. Cox & NgNetworking72 Echo Server: echo The server uses RIO to read and echo text lines until EOF (end-of-file) is encountered  EOF notification caused by client calling close(clientfd)  IMPORTANT: EOF is a condition, not a particular data byte void echo(int connfd) { size_t n; char buf[MAXLINE]; rio_t rio; Rio_readinitb(&rio, connfd); while((n = Rio_readlineb(&rio, buf, MAXLINE)) != 0) { printf("server received %d bytes\n", n); Rio_writen(connfd, buf, n); }

73. Cox & NgNetworking73 Testing Servers Using telnet The telnet program is invaluable for testing servers that transmit ASCII strings over Internet connections  Our simple echo server  Web servers  Mail servers Usage:  unix> telnet  Creates a connection with a server running on and listening on port

74. Cox & NgNetworking74 Testing the Echo Server With telnet bass> echoserver 5000 server established connection with KITTYHAWK.CMCL (128.2.194.242) server received 5 bytes: 123 server established connection with KITTYHAWK.CMCL (128.2.194.242) server received 8 bytes: 456789 kittyhawk> telnet bass 5000 Trying 128.2.222.85... Connected to BASS.CMCL.CS.CMU.EDU. Escape character is '^]'. 123 Connection closed by foreign host. kittyhawk> telnet bass 5000 Trying 128.2.222.85... Connected to BASS.CMCL.CS.CMU.EDU. Escape character is '^]'. 456789 Connection closed by foreign host. kittyhawk>

75. Cox & NgNetworking75 Running the Echo Client and Server bass> echoserver 5000 server established connection with KITTYHAWK.CMCL (128.2.194.242) server received 4 bytes: 123 server established connection with KITTYHAWK.CMCL (128.2.194.242) server received 7 bytes: 456789... kittyhawk> echoclient bass 5000 Please enter msg: 123 Echo from server: 123 kittyhawk> echoclient bass 5000 Please enter msg: 456789 Echo from server: 456789 kittyhawk>

76. Cox & NgNetworking76 For More Information W. Richard Stevens, “Unix Network Programming: Networking APIs: Sockets and XTI”, Volume 1, Second Edition, Prentice Hall, 1998.  THE network programming bible Complete versions of the echo client and server are developed in the text  Available on the course web site  You should compile and run them for yourselves to see how they work  Feel free to borrow any of this code

77. Cox & NgNetworking77 Web History 1945:  Vannevar Bush, “As we may think”, Atlantic Monthly, July, 1945 Describes the idea of a distributed hypertext system A “memex” that mimics the “web of trails” in our minds 1989:  Tim Berners-Lee (CERN) writes internal proposal to develop a distributed hypertext system Connects “a web of notes with links” Intended to help CERN physicists in large projects share and manage information 1990:  Tim Berners-Lee writes a graphical browser for Next machines

78. Cox & NgNetworking78 Web History (cont) 1992  NCSA server released  26 WWW servers worldwide 1993  Marc Andreessen releases first version of NCSA Mosaic browser  Mosaic version released for (Windows, Mac, Unix)  Web (port 80) traffic at 1% of NSFNET backbone traffic  Over 200 WWW servers worldwide 1994  Andreessen and colleagues leave NCSA to form "Mosaic Communications Corp" (became Netscape, then part of AOL)

79. Cox & NgNetworking79 Internet Hosts Source: www.isc.org

80. Cox & NgNetworking80 Web Servers Clients and servers communicate using the HyperText Transfer Protocol (HTTP)  Client and server establish TCP connection  Client requests content  Server responds with requested content  Client and server (may) close connection Current version is HTTP/1.1  RFC 2616, June, 1999 Web server HTTP request HTTP response (content) Web client (browser)

81. Cox & NgNetworking81 Web Content Web servers return content to clients  content: a sequence of bytes with an associated MIME (Multipurpose Internet Mail Extensions) type Example MIME types  text/html HTML document  text/plain Unformatted text  application/postscript Postcript document  image/gif Binary image (GIF format)  image/jpeg Binary image (JPEG format)

82. Cox & NgNetworking82 Static and Dynamic Content The content returned in HTTP responses can be either static or dynamic  Static content: content stored in files and retrieved in response to an HTTP request Examples: HTML files, images, audio clips  Dynamic content: content produced on-the-fly in response to an HTTP request Example: content produced by a program executed by the server on behalf of the client (i.e., search results) Bottom line: All Web content is associated with a file that is managed by the server

83. Cox & NgNetworking83 URLs Each file managed by a server has a unique name called a URL (Universal Resource Locator) URLs for static content:  http://www.rice.edu:80/index.html  http://www.rice.edu/index.html  http://www.rice.edu Identifies a file called index.html, managed by a Web server at www.rice.edu that is listening on port 80 URLs for dynamic content:  http://www.cs.cmu.edu:8000/cgi-bin/adder?15000&213 Identifies an executable file called adder, managed by a Web server at www.cs.cmu.edu that is listening on port 8000, that should be called with two argument strings: 15000 and 213

84. Cox & NgNetworking84 How Clients and Servers Use URLs Example URL: http://www.aol.com:80/index.html Clients use prefix ( http://www.aol.com:80 ) to infer:  What kind of server to contact ( http (Web) server)  Where the server is ( www.aol.com )  What port the server is listening on ( 80 ) Servers use suffix ( /index.html ) to:  Determine if request is for static or dynamic content No hard and fast rules for this Convention: executables reside in cgi-bin directory  Find file on file system Initial “/” in suffix denotes home directory for requested content Minimal suffix is “/”, which servers expand to some default home page (e.g., index.html)

85. Cox & NgNetworking85 Anatomy of an HTTP Transaction unix> telnet www.google.com 80 Client: open connection to server Trying 209.85.164.104... Telnet prints 3 lines to the terminal Connected to www.l.google.com. Escape character is '^]'. GET / HTTP/1.1 Client: request line host: www.google.com Client: required HTTP/1.1 HOST header Client: empty line terminates headers HTTP/1.1 200 OK Server: response line Cache-Control: private Server: followed by six response headers Content-Type: text/html; charset=ISO-8859-1 Set-Cookie: PREF=ID= Server: gws Transfer-Encoding: chunked Date: Tue, 13 Nov 2007 17:25:04 GMT Server: empty line (“ \r\n ”) terminates hdrs Server: first HTML line in response body Server: HTML content not shown. Server: last HTML line in response body Connection closed by foreign host. Server: closes connection unix> Client: closes connection and terminates

86. Cox & NgNetworking86 HTTP Requests HTTP request is a request line, followed by zero or more request headers Request line:  is either GET, POST, OPTIONS, HEAD, PUT, DELETE, or TRACE  is typically URL for proxies, URL suffix for servers  is HTTP version of request (HTTP/1.0 or HTTP/1.1)

87. Cox & NgNetworking87 HTTP Requests (cont) HTTP methods:  GET: Retrieve static or dynamic content Arguments for dynamic content are in URI Workhorse method (99% of requests)  POST: Retrieve dynamic content Arguments for dynamic content are in the request body  OPTIONS: Get server or file attributes  HEAD: Like GET but no data in response body  PUT: Write a file to the server!  DELETE: Delete a file on the server!  TRACE: Echo request in response body Useful for debugging

88. Cox & NgNetworking88 HTTP Requests (cont) Request headers: :  Provide additional information to the server Major differences between HTTP/1.1 and HTTP/1.0  HTTP/1.0 uses a new connection for each transaction  HTTP/1.1 also supports persistent connections Multiple transactions over the same connection Connection: Keep-Alive  HTTP/1.1 requires HOST header Host: www.yahoo.com  HTTP/1.1 adds additional support for caching

89. Cox & NgNetworking89 HTTP Responses HTTP response is a response line followed by zero or more response headers Response line:  is HTTP version of the response  is numeric status  is corresponding English text 200 OKRequest was handled without error 403ForbiddenServer lacks permission to access file 404Not foundServer couldn’t find the file Response headers: :  Provide additional information about response  Content-Type : MIME type of content in response body  Content-Length : Length of content in response body

90. Cox & NgNetworking90 GET Request to Apache Server From Firefox Browser GET /~comp221/proxytest.html HTTP/1.1 Host: www.owlnet.rice.eduwww.owlnet.rice.edu User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.8) Gecko/20071022 Ubuntu/7.10 (gutsy) Firefox/2.0.0.8 Accept: text/xml,application/xml,application/xhtml+xml,text/html; q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5 Accept-Language: en-us,en;q=0.5 Accept-Encoding: gzip,deflate Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7 Connection: keep-alive Keep-Alive: 300 CRLF (\r\n)

91. Cox & NgNetworking91 GET Response From Apache Server HTTP/1.1 200 OK Date: Tue, 13 Nov 2007 18:09:01 GMT Server: Apache/1.3.26 (Unix) mod_ssl/2.8.9 OpenSSL/0.9.6e Last-Modified: Tue, 13 Nov 2007 18:08:44 GMT ETag: “ac330311-df-4739e82c" Accept-Ranges: bytes Content-Length: 212 Connection: close Content-Type: text/html CRLF COMP221 Web Proxy Test Page This page is a simple text-only HTML file that can be used to test your web proxy software.

92. Cox & NgNetworking92 Serving Dynamic Content Client sends request to server Server parses request URI to determine if request is for static or dynamic content  In the “old” days, this was as simple as the string “/cgi-bin” starting the URL  Web servers are far more flexible and configurable now ClientServer GET /cgi-bin/env.pl HTTP/1.1

93. Cox & NgNetworking93 Serving Dynamic Content (cont) The server creates a child process and runs the program identified by the URI in that process ClientServer env.pl fork/exec

94. Cox & NgNetworking94 Serving Dynamic Content (cont) The child runs and generates the dynamic content The server captures the content of the child and forwards it without modification to the client ClientServer env.pl Content

95. Cox & NgNetworking95 Issues in Serving Dynamic Content How does the client pass program arguments to the server? How does the server pass these arguments to the child? How does the server pass other info relevant to the request to the child? How does the server capture the content produced by the child? These issues are addressed by the Common Gateway Interface (CGI) specification ClientServer Content Request Create env.pl

96. Cox & NgNetworking96 CGI Because the children are written according to the CGI spec, they are often called CGI programs Because many CGI programs are written in Perl, they are often called CGI scripts However, CGI really defines a simple standard for transferring information between the client (browser), the server, and the child process

97. Cox & NgNetworking97 add.com: THE Internet addition portal! Ever need to add two numbers together and you just can’t find your calculator? Try the addition service at “add.com: THE Internet addition portal!”  Takes as input the two numbers you want to add together  Returns their sum in a tasteful personalized message

98. Cox & NgNetworking98 The add.com Experience input URL Output page hostportCGI programargs

99. Cox & NgNetworking99 Serving Dynamic Content With GET Question: How does the client pass arguments to the server? Answer: The arguments are appended to the URI Can be encoded directly in a URL typed to a browser or a URL in an HTML link  http://add.com/cgi-bin/adder?1&2  adder is the CGI program on the server that will do the addition  argument list starts with “?”  arguments separated by “&”  spaces represented by “+” or “%20” Can also be generated by an HTML form

100. Cox & NgNetworking100 Serving Dynamic Content With GET URL:  http://add.com/cgi-bin/adder?1&2 Result displayed on browser: Welcome to add.com: THE Internet addition portal. The answer is: 1 + 2 = 3 Thanks for visiting! Tell your friends.

101. Cox & NgNetworking101 Serving Dynamic Content With GET Question: How does the server pass these arguments to the child? Answer: In environment variable QUERY_STRING  A single string containing everything after the “?”  For add.com: QUERY_STRING = “1&2” /* child code that accesses the argument list */ if ((buf = getenv("QUERY_STRING")) == NULL) { exit(1); } /* extract arg1 and arg2 from buf and convert */... n1 = atoi(arg1); n2 = atoi(arg2);

102. Cox & NgNetworking102 Serving Dynamic Content With GET Question: How does the server pass other info relevant to the request to the child? Answer: In a collection of environment variables defined by the CGI spec

103. Cox & NgNetworking103 Some CGI Environment Variables General  SERVER_SOFTWARE  SERVER_NAME  GATEWAY_INTERFACE (CGI version) Request-specific  SERVER_PORT  REQUEST_METHOD ( GET, POST, etc)  QUERY_STRING (contains GET arguments)  REMOTE_HOST (domain name of client)  REMOTE_ADDR (IP address of client)  CONTENT_TYPE (for POST, type of data in message body, e.g., text/html )  CONTENT_LENGTH (length in bytes)

104. Cox & NgNetworking104 Some CGI Environment Variables In addition, the value of each header of type type received from the client is placed in environment variable HTTP_type  Examples: HTTP_ACCEPT HTTP_HOST HTTP_USER_AGENT (any “-” is changed to “_”)

105. Cox & NgNetworking105 Serving Dynamic Content With GET Question: How does the server capture the content produced by the child? Answer: The child generates its output on stdout  Server uses dup2 to redirect stdout to its connected socket  Notice that only the child knows the type and size of the content, so the child (not the server) must generate the corresponding headers /* child generates the result string */ sprintf(content, "Welcome to add.com: THE Internet addition portal\ The answer is: %d + %d = %d\ Thanks for visiting!\r\n", n1, n2, n1+n2); /* child generates the headers and dynamic content */ printf("Content-length: %d\r\n", strlen(content)); printf("Content-type: text/html\r\n"); printf("\r\n"); printf("%s", content);

106. Cox & NgNetworking106 Serving Dynamic Content With GET bass>./tiny 8000 GET /cgi-bin/adder?1&2 HTTP/1.1 Host: bass.cmcl.cs.cmu.edu:8000 kittyhawk> telnet bass 8000 Trying 128.2.222.85... Connected to BASS.CMCL.CS.CMU.EDU. Escape character is '^]'. GET /cgi-bin/adder?1&2 HTTP/1.1 Host: bass.cmcl.cs.cmu.edu:8000 HTTP/1.1 200 OK Server: Tiny Web Server Content-length: 102 Content-type: text/html Welcome to add.com: THE Internet addition portal. The answer is: 1 + 2 = 3 Thanks for visiting! Connection closed by foreign host. kittyhawk> HTTP request received by Tiny Web server HTTP request sent by client HTTP response generated by the server HTTP response generated by the CGI program

107. Cox & NgNetworking107 Proxies A proxy is an intermediary between a client and an origin server  To the client, the proxy acts like a server  To the server, the proxy acts like a client ClientProxy Origin Server HTTP request HTTP response

108. Cox & NgNetworking108 Why Proxies? Can perform useful functions as requests and responses pass through  Examples: Caching, logging, anonymization Client A Proxy cache Origin Server Request foo.html foo.html Client B Request foo.html foo.html Fast inexpensive local network Slower more expensive global network

109. Cox & NgNetworking109 For More Information Study the Tiny Web server described in your text  Tiny is a sequential Web server  Serves static and dynamic content to real browsers text files, HTML files, GIF and JPEG images  220 lines of commented C code  Also comes with an implementation of the CGI script for the add.com addition portal

110. Cox & NgNetworking110 Next Time Concurrency