What is a socket?An interface between application and

networkThe application creates a socketThe socket type dictates the style of

communication reliable vs. best effort connection-oriented vs. connectionless

Once configured the application canpass data to the socket for network

transmissionreceive data from the socket (transmitted

through the network by some other host)

What is Socket ?Endpoint of any connectionTwo Types :

TCPUDP

Identified by Two valuesAn IP AddressA Port Number

Two essential types of sockets

SOCK_STREAM

reliable delivery in-order guaranteedconnection-orientedbidirectional

SOCK_DGRAM

unreliable deliveryno order guaranteesno notion of “connection” –

app indicates dest. for each packet

can send or receive

4

App

socket3 2 1

Dest.

App

socket3 2 1

D1

D3

D2

Socket Functionssocket()

bind()

listen()

accept()socket()

connect()

write()

read()

close() read()

close()

read()

write()

Connection establishment(TCP three-way handshake)

Data (request)

Data (reply)

End-of-file notification

Process request

Blocks until connection from client

CLIENT SIDE

SERVER SIDE

Well known port

Running an App

Socket Creation in C: socketint s = socket(domain, type, protocol);

s: socket descriptor, an integer (like a file-handle)

domain: integer, communication domain e.g., AF_INET (IPv4 protocol) – typically used

type: communication type SOCK_STREAM: reliable, 2-way, connection-based

service SOCK_DGRAM: unreliable, connectionless,

Socket Creation in C: socketprotocol: specifies protocol (see file /etc/protocols

for a list of options) – usually set to 0 to select the system’s default

for the given combination of family and type.

NOTE: socket call does not specify where data will be coming from, nor where it will be going to – it just creates the interface!

Protocol family constants

Family Description

AF_INET IPv4 protocol

AF_INET6 IPv6 protocol

AF_LOCAL UNIX DOMAIN PROTOCOL

AF_ROUTE Routing socket

AF_KEY Key socket

Type of socketType DescriptionSOCK_STREAM STREAM socket

SOCK_DGRAM Datagram socket

SOCK_SEQPACKET

Sequenced packet socket

SOCK_RAW Raw socket

Protocol of socketprotocol description

IPPROTO_TCP TCP transport protocol

IPPROTO_UDP UDP transport protocol

IPPROTO_SCTP SCTP transport protocol

Addresses, Ports and SocketsLike apartments and mailboxes

You are the applicationYour apartment building address is the addressYour mailbox is the portThe post-office is the networkThe socket is the key that gives you access to the right

mailbox (one difference: assume outgoing mail is placed by you in your mailbox)

IPv4 Socket Address StructureStruct in_addr{ in_addr_t s_addr; /*32bit IPv4 address*/ }; /*network byte ordered*/

struct sockaddr_in { uint8_t sin_len; /* length of structure(16) */ sa_family_t sin_family; /* AF_INET */ in_port_t sin_port; /* 16bit TCP or UDP port

number */ /*network byte ordered*/ struct in_addr sin_addr; /* 32bit IPv4 address */ /*network byte ordered*/ char sin_zero[8]; /* unused */ }; /* included in <netinet/in.h> */

Socket Address StructureLength field simplifies the handling of

variable-length socket address structures.Used with routing socket.In_addr_t datatype must be an unsigned

integer type of at least 32 bits.

DatatypeDatatype Description Header

Int8_t Signed 8-bit integer <sys/types.h>

Uint8_t unsigned 8-bit integer <sys/types.h>

Int16_t Signed 16-bit integer <sys/types.h>

Uint16_t unsigned 16-bit integer <sys/types.h>

Int32_t Signed 32-bit integer <sys/types.h>

Uint32_t unsigned 32-bit integer <sys/types.h>

Sa_family_t

Add. Family of socket add struct

<sys/socket.h>

Socklen_t Length of socket add struct

<sys/socket.h>

In_addr_t IPV4 address, uint32_t <netinet/in.h>

In_port_t TCP or UDP port, uint16_t

<netinet/in.h>

Generic Socket Address structure

A Socket address structure must be passed by reference

socket function that takes one of these pointers as an argument must deal with socket address structures from any of the supported protocol families.

How to declare the type of pointerSoln : void *Define Generic socket address structure<sys/socket.h>

Generic Socket Address structure

Struct sockaddr

{

uint8_t sa_len;

sa_family_t sa_family;

char sa_data[14];/* protocol specific address*/

};

From an application programmer's point of view, the only use of these generic socket address structures is to cast pointers to protocol-specific structures.

IPv6 Socket Address StructureStruct in6_addr{ uint8_t s6_addr[16]; /*128bit IPv6 address*/ }; /*network byte ordered*/#define SIN6_LEN /* required for compile-time tests */struct sockaddr_in6 { uint8_t sin6_len; /* length of structure(24) */ sa_family_t sin6_family; /* AF_INET6*/ in_port_t sin6_port; /* Transport layer port# */ /*network byte ordered*/ uint32_t sin6_flowinfo; /* priority & flow label */ /*network byte ordered*/ struct in6_addr sin6_addr; /* IPv6 address */ /*network byte ordered*/}; /* included in <netinet/in.h> */

New Generic Socket Address structure

Struct sockaddr {

uint8_t sa_len;sa_family_t sa_family;/* implementation dependent elements to provide1. alignment2. enough storage to hold any type of socket

address that the system supports */ };

sockaddr_storage different from struct sockaddr in two ways:If any socket address structures that the

system supports have alignment requirements, the sockaddr_storage provides the strictest alignment requirement.

The sockaddr_storage is large enough to contain any socket address structure that the system supports.

Comparison of socket address structure

Value-result ArgumentsAccept , recvfrom , getpeername pass socket

address structure from kernel to the process.

Size changes from an int to pointer to an integer is because the size is both a value when the function is called and Result when function returns.

connect (sockfd, (SA *) &serv, sizeof(serv));

getpeername(unixfd, (SA *) &cli, &len);

23

Address and port byte-orderingAddress and port are stored as

integersu_short sin_port; (16 bit) in_addr sin_addr; (32 bit) Problem:

different machines / OS’s use different word orderings• little-endian: lower bytes first• big-endian: higher bytes first

these machines may communicate with one another over the network

128.119.40.12

128

119

40 12

12.40.119.128

128

119

40 12

Big-Endianmachine Little-Endian

machine

WRONG!!

!

Solution: Network Byte-OrderingDefs:

Host Byte-Ordering: the byte ordering used by a host (big or little)

Network Byte-Ordering: the byte ordering used by the network – always big-endian

Any words sent through the network should be converted to Network Byte-Order prior to transmission (and back to Host Byte-Order once received)

Byte Ordering FunctionsprogramTwo types of Byte Ordering

Little-endian Byte Ordering

Big-endian Byte Ordering

MSB LSB

Address A+1

Address A

MSB LSB

Address A Address A+1

determine host byte orderint main(int argc, char **argv){

union { short s; char c[sizeof(short)]; } un;

un.s = 0x0102;printf("%s: ", CPU_VENDOR_OS);if (sizeof(short) == 2)

{ if (un.c[0] == 1 && un.c[1] == 2) printf("big-endian\n"); else if (un.c[0] == 2 && un.c[1] == 1) printf("little-endian\n"); else printf("unknown\n");

} else printf("sizeof(short) = %d\n", sizeof(short)); exit(0);}

Byte Ordering Functions<netinet/in.h>uint16_t htons(uint16_t host16val); -- convert 16-bit

value from host to network order, used for the port number.

uint32_t htonl(uint32_t host32val); -- convert 32-bit value from host to network order, used for the Ipv4

address.uint16_t ntohs(uint16_t network16val); -- convert 16-bit

value from network to host order.uint32_t ntohl(uint32_t network32val); -- convert 32-bit

value from network to host order.

Byte Ordering Functions#include<strings.h>

Void bzero(void *dest, size_t nbytes); -- SETS THE SPECIFIED NO. OF BYTES TO O IN THE DESTINATION.

Void bcopy(const void *src, void *dest, size_t nbytes); -- copies nbytes of src to dest.

int bcmp(const void *ptrl, const void *ptr2, size_t nbytes);

-- compare nbytes of the two strings, Returns 0 if they match, >0 if ptr1 >ptr2, < 0 if ptr1 <

ptr2.

void *memset(void *dest, int c, size_t nbytes); ---- writes value c in the destination

----bytes of dest. Returns dest.void *memcpy(void *dest, const void *src,

size_t nbytes);---- copies nbytes of src to dest. Returns dest.

int memcmp(const void *ptrl, const void *ptr2, size_t nbutes);

---- compare nbytes of the two strings, Returns 0 if

they match, >0 if ptr1 >ptr2, < 0 if ptr1 < ptr2.

Inet_aton, inet_addr, inet_ntoaAn internet address is written as: “192. 43.

234.1”, saved as a character string. The functions require their number as a 32-bit binary value.

<arpa/inet.h>

int inet_aton(const char *strptr, struct in_addr * addrptr);

-- returns 1, 0 on error.

Converts from a dotted-decimal string to a network

address.

Inet_aton, inet_addr, inet_ntoaIn_addr_t inet_addr(const char *strptr);

-- Converts from a dotted-decimal string to 32-bit interger as return value.--INADDR_NONE IF ERRORs

char *inet_ntoa (struct in_addr addrptr);

-- Returns a pointer to a dotted-decimal string, given a valid network address.

Inet_pton functions

#include<arpa/inet.h>

Convert strptr ( which is in presentation form ) to numeric form and store it into addrptr.

int inet_pton(int family, const char *strptr, void *addrptr);-- returns 1 if OK, 0 if invalid input format, -1 on error.

inet_ntop functionsConst char *inet_ntop (int family, const void

*addrptr, char *strptr, size_t len);-- returns pointer to result if OK, NULL on error.

--len argument is the size of the destination, to prevent the function from overflowing the caller’s buffer

Two functions supports both IPv4 and IPv6 protocol

Sock_ntop function

A basic problem with inet_ntop is that it requires the caller to pass a pointer to a binary address.

This address is normally contained in a socket address structure,

Requiring the caller to know the format of the structure and the address family.

Sock_ntop functionstruct sockaddr_in addr;

inet_ntop(AF_INET, &addr.sin_addr, str, sizeof(str)); For IPV6 struct sockaddr_in6 addr6;

inet_ntop(AF_INET6, &addr6.sin6_addr, str, sizeof(str));

sock_ntop that takes a pointer to a socket address structure, looks inside the structure, and calls the appropriate function to return the presentation format of the address.

Sock_ntop functionchar *sock_ntop(const struct sockaddr *sockaddr,

socklen_t addrlen);Returns: non-null pointer if OK, NULL on error

sockaddr points to a socket address structure whose length is addrlen.

The function uses its own static buffer to hold the result and a pointer to this buffer is the return value.

sock_ntop and related functions.

int sock_bind_wild(int sockfd, int family);

Returns: 0 if OK, -1 on error

sock_bind_wild binds the wildcard address and an ephemeral port to a socket.

sock_ntop and related functions.

sock_cmp_addr compares the address portion of two socket address structures

int sock_cmp_addr(const struct sockaddr *sockaddr1,const struct sockaddr *sockaddr2, socklen_t addrlen);

Returns: 0 if addresses are of the same family and ports are equal, else nonzero

sock_ntop and related functionsint sock_cmp_port(const struct sockaddr

*sockaddr1,const struct sockaddr *sockaddr2, socklen_t addrlen);

Returns: 0 if addresses are of the same family and ports are equal, else nonzero

sock_cmp_port compares the port number of two socket address structures.

sock_ntop and related functions sock_get_port returns just the port number

int sock_get_port(const struct sockaddr *sockaddr, socklen_t addrlen);

Returns: non-negative port number for IPv4 or IPv6 address, else -1

sock_ntop and related functions

char *sock_ntop_host(const struct sockaddr *sockaddr, socklen_t addrlen);

Returns: non-null pointer if OK, NULL on error

sock_ntop_host converts just the host portion of a socket address structure to presentation format (not the port number).

sock_ntop and related functions

void sock_set_addr(const struct sockaddr *sockaddr, socklen_t addrlen, void *ptr);

sock_set_addr sets just the address portion of a socket address structure to the value pointed to by ptr

sock_ntop and related functionsvoid sock_set_port(const struct sockaddr

*sockaddr, socklen_t addrlen, int port)

sock_set_port sets just the port number of a socket address structure

void sock_set_wild(struct sockaddr *sockaddr, socklen_t addrlen);

sock_set_wild sets the address portion of a socket address structure to the wildcard

readn, writen, and readline Functions A read or write on a stream socket might input or

output fewer bytes than requested.Buffer limits might be reached for the socket in

the kernel.#include "unp.h“ssize_t readn(int filedes, void *buff, size_t nbytes);ssize_t writen(int filedes, const void *buff, size_t

nbytes);ssize_t readline(int filedes, void *buff, size_t

maxlen);All return: number of bytes read or written, –1 on

error

readn function: Read n bytes from a descriptor. #include "unp.h" ssize_t readn(int fd, void *vptr, size_t n) { size_t nleft; ssize_t nread; char *ptr; ptr =

vptr; nleft = n; while (nleft > 0) { if ( (nread = read(fd, ptr, nleft)) < 0) { if (errno == EINTR) nread = 0; /* and call read() again

*/ else return (-1); } else if (nread == 0) break; /* EOF */ nleft -= nread; ptr += nread; } return (n - nleft); /* return >= 0 */ }

writen function: Write n bytes to a descriptor ssize_t writen(int fd, const void *vptr, size_t n) { size_t nleft; ssize_t nwritten; const char *ptr; ptr = vptr; nleft = n; while (nleft > 0) { if ( (nwritten = write(fd, ptr, nleft)) <= 0) { if (nwritten < 0 && errno == EINTR) nwritten = 0; /* and call write() again */ else return (-1); /* error */ } nleft -= nwritten; ptr += nwritten; } return (n); }

readline function: Read a text line from a

descriptor, one byte at a time. ssize_t readline(int fd, void *vptr, size_t maxlen) { ssize_t n, rc; char c, *ptr; ptr = vptr; for (n = 1; n < maxlen; n++) { again : if ( (rc = read(fd, &c, 1)) == 1) { *ptr++ = c; if (c == '\n') break; /* newline is stored, like fgets() */ } else if (rc == 0)

{ *ptr = 0; return (n - 1); /* EOF, n - 1 bytes were read */

} else { if (errno == EINTR) goto again;

return (-1); /* error, errno set by read() */ } } *ptr = 0; /* null terminate like fgets() */ return (n); }

Example – Daytime Server/Client

TCP

MAC driver

Daytime client

IP

Network

TCP

Daytime server

IP

MAC driver

Application protocol

TCP protocol

IP protocol

MAC-level protocol

Actual data flow

Socket API Socket API

MAC = media access control

Files to be createdClient Server

Client ServerClient source file - client.cServer source file - server.cHeader file - header.h

STRING

Running an ApplicationCompile and run server file ( you can run it as a

background process too)e.g. $cc server.c

$./a.out 2345 &Compile and Run client file

e.g. $cc client.c$./a.out 172.16.2.2 2345

Thu March 05 15:50:00 2009

Back

abstractSocket functionconnect functionbind functionlisten functionaccept functionfork and exec function concurrent serverclose functiongetsockname and getpeername function

Socket Functionssocket()

bind()

listen()

accept()socket()

connect()

write()

read()

close() read()

close()

read()

write()

Connection establishment(TCP three-way handshake)

Data (request)

Data (reply)

End-of-file notification

Process request

Blocks until connection from client

CLIENT SIDE

SERVER SIDE

Well known port

Running an App

#include <sys/socket.h>int socket(int family, int type, int protocol); returns:nonnegative descriptor if OK,

-1 on error

==>Normaly the protocol argument to the socket function is set to 0 exept for raw socket.

connect Function #include <sys/socket.h>

int connect(int sockfd, const struct sockaddr *servaddr, socklen_t addrlen);

Returns : 0 if successful connect, -1 otherwisesockfd: integer, socket to be used in

connection struct sockaddr: address of passive

participant integer, sizeof(struct)

(If connect fails, the SYN_SENT socket is no longer useable.)

Connection Setup (SOCK_STREAM) A connection occurs between two kinds of

participantspassive: waits for an active participant to

request connectionactive: initiates connection request to passive

side

Once connection is established, passive and active participants are “similar”both can send & receive dataeither can terminate the connection

Connect functionReturn error

ETIMEOUT : no response from serverRST : server process is not runningEHOSTUNREACH : client’s SYN unreachable

from some intermediate router.

Error Return by ConnectIf the client TCP receives no response to

its SYN segment, ETIMEDOUT is returned.

If the server’s response to the client’s SYN is a reset (RST), this indicates that no process is waiting for connections on the server host at the port specified.Hard errorECONNREFUSED is returned to the client as soon as

the RST is received.

Error Return by ConnectThree conditions that generate an RST are:

When a SYN arrives for a port that has no listening server

When TCP wants to abort an existing connectionWhen TCP receives a segment for a connection

that does not exist.

Error Return by ConnectICMP destination unreachable received in

response to client TCP’s SYN (maybe due to transient routing problem), resend SYN timeout after 75 sec, returns EHOSTUNREACH or ENETUNREACH

connect Function: Three-Way Handshake

• No bind before connect :The kernel chooses the source IP, if necessary, and an ephemeral port (for the client).

Hard error: RST received in response to client TCP’s SYN (server not running) returns ECONNREFUSED

Soft error: 1. no response to client TCP’s SYN, resend SYN, timeout after 75 sec (in 4.4BSD), returns ETIMEOUT

2. ICMP destination unreachable received in response to client TCP’s SYN (maybe due to transient routing problem), retx SYN, timeout after 75 sec, returns EHOSTUNREACH)

The bind functionAssigns a local protocol address to a

socket.int status = bind(int sockid, const struct

sockaddr &myaddr, socklen_t addrlen);status: error status, = -1 if bind failed, 0 if OK.sockid: integer, socket descriptormyaddr: struct sockaddr, the (IP) address and

port of the machine (address usually set to INADDR_ANY – chooses a local address)

addrlen: the size (in bytes) of the addrport structure

bind FunctionUsually servers bind themselves to their well-

known ports.RPC servers let kernel choose ephemeral

ports which are then registered with the RPC port mapper.

Normally, TCP client does not bind an IP address to its socket.

If a TCP server does not bind an IP address to its socket, the kernel uses the destination IP address of the client’s SYN as the server’s source IP address

bind Function

Table summarizes the values to which we set sin_addr and sin_port depending on the desired result.

Process specifiesIP address port ResultWildcard 0 kernel chooses IP addr and portwildcard nonzero kernel chooses IP addr, process specifies portlocal IP addr 0 kernel chooses port, process specifies IP addrlocal IP addr nonzero process specifies IP addr and port

bind FunctionWildcard address is specified by the constant

INADDR_ANY whose value is normally 0.To obtain value of the ephemeral port

assigned by the kernel, we must call getsockname to return the protocol address.

listen function#include <sys/socket.h>int listen(int sockfd, int backlog);

Returns:0 if OK, -1 on error

==>This function is called only by a TCP serverThe listen function converts an unconnected socket into a

passive socket, indicating that the kernel should accept incoming connection requests directed to this socket.

backlog =>specify the maximum number of connections that the kernel should queue for this socket.

If the queues are full when client SYN arrives, TCP server ignore the SYN, it does not send RST.

listen function

●An incomplete connection queue, which contains an entry for each SYN that has arrived from a client for which the server is awaiting completion of the TCP three-way handshake

●A completed connection queue, which contains an entry for each client with whom the TCP three-way handshake has completed

Backlog argument to the listen function has historically specified the maximum value for the sum of both queues

listen Function

Backlog argument to listen function has specified the maximum value for the sum of both queues

Berkeley derived : multipied by 1.5Do not specify backlog of 0What value should the application specify?Allow Command line or an environment

variable to override default.

Listen that allows an environment var to specify backlog

Void Listen(int fd, int backlog){ char *ptr; if ((ptr=getenv(“LISTENQ”))!=NULL)

backlog=atoi(ptr); if (listen(fd,backlog)<0) printf(“listen error”);}

listen Function

If the queues are full when a client SYN arrives, TCP ignores the arriving SYN : it does not send an RST.

Data that arrives after the three way handshake completes, but before the server call accept, should be queued by the server TCP, up to the size of the connected socket’s receive buffer.

accept function#include <sys/socket.h>int accept(int sockfd, struct sockaddr *cliaddr, socklen_t *addrlen);

Returns:nonnegative descriptor if OK, -1 on error

=> return the next completed connection from the front of the completed connection queue.

If queue is empty, the process is put to sleep.

Return three values:1. integer return code 2. protocol address of the client process3. size of this address :

This integer value contains the actual number of bytes

stored by the kernel in the socket address structure.

If we are not interested in having the protocol address of the client returned, we set both cliaddr and addrlen to null pointers.

fork and exec function#include <unistd.h>

pid_t fork(void); Returns: 0 in child, process ID of child in parent, -1 on error

#include <unistd.h>int execl(const char *pathname, const char *arg(), …/*(char *)

0*/);int execv(const char *pathname, char *const argv[]);int execle(const char *pathname, const char *arg());int execve(const char *pathname, char *const argv[], char

*const envp[]);int execlp(const char *filename, const char *arg());int execvp(const char *filename, char *const argv[]); All six return: -1 on error, no return on

success

A process makes a copy of itself so that one copy can handle one operation while the other copy does another task.

A process wants to execute another program. Since the only way to create a new process is by calling fork, the process first calls fork to make a copy of itself, and then one of the copies (typically the child process) calls exec to replace itself with the new program.

The differences in the six exec functions are: (a) whether the program file to execute is

specified by a filename or a pathname; (b) whether the arguments to the new program

are listed one by one or referenced through an array of pointers; and

(c) Whether the environment of the calling process is passed to the new program or whether a new environment is specified.

Concurrent serverpid_t pidint listenfd, connfd;listenfd = Socket(...);//fill in sockaddr_in{} with server’s well-known portBind(listenfd, LISTENQ);for(;;){ connfd = Accept(listenfd, ...); if( (pid = Fork()) == 0) { Close(listenfd); /* child closes listening socket */ doit(connfd); //process the request Close(); //done with this client exit(0); //child terminate } Close(connfd); // parent close connected socket} Figure 4.13 Outline for typical concurrent server

closeWhen finished using a socket, the socket

should be closed:status = close(s);

status: 0 if successful, -1 if errors: the file descriptor (socket being closed)

Closing a socketcloses a connection (for SOCK_STREAM)frees up the port used by the socket

Return the address family of socket#include "unp.h" int sockfd_to_family(int sockfd) { struct sockaddr_storage ss; socklen_t len; len = sizeof(ss); if (getsockname(sockfd, (SA *) &ss, &len) < 0) return (-1); return (ss.ss_family); }

getsockname and getpeername function#include<sys/socket.h>int getsockname(int sockfd, struct sockaddr

*localaddr, socklen_t *addrlen);int getpeername(int sockfd, struct sockaddr

*peeraddr, socklen_t *addrlen);

both return : 0 if OK, -1 on error=>getsockname : return local address associated with a socket getpeername : foreign protocol address associated with a socket

getsockname and getpeername

This function takes the following three input arguments:

1. The sockets to query for the socket address.2. The pointer to the receiving buffer (argument

name).3. Pointer to the maximum length variable. This

variable provides the maximum length in bytes that can be received in the buffer (argument namelen).

Reasons:TCP client that does not call bind,

getsockname returns the local protocol address

After calling bind with a port no. of 0, getsockname returns the local port number

It can be called to obtain the address family of a socket.

Server binds to wildcard IP address,getsockname used to obtain the local protocol address

Only way to obtain the identity of client is to call getpeername

Value-result Argumentswhen a socket address structure is passed to any

socket function, it is always passed by reference. That is, a pointer to the structure is passed. The length of the structure is also passed as an

argument. But the way in which the length is passed depends on which direction the structure is being passed: from the process to the kernel, or vice versa.

Bind, connect, and sendto pass socket address structure from the process to kernel

Files to be createdClient Server

Client ServerClient source file - client.cServer source file - server.cHeader file - header.h

STRING

TCP Echo Client/Server The client reads a line of text from its

standard input and writes the line to the server.

The server reads the line from its network input and echoes the line back to the client.

The client reads the echoed line and prints it on its standard output.

TCP Echo Server: main Function Create socket, bind server's well-known

portWait for client connection to complete Concurrent server H:\echo_server.cHeader file : unp.h

TCP echo client. Create socket, fill in Internet socket

address structureConnect to serverG:\echo_client.c.txt

TCP Echo Client: str_cli Function Read a line, write to server Read echoed line from server, write to

standard outputReturn to main