I've a data acquisition module from which I would like to collect data from the Ethernet port. I'm getting there in steps, currently I would like to just connect to server from a client. I've used Beej's guide to get the basic C code. But I just keep getting this connect error connect: Connection refused.
This is what I do:
The network IP mentioned here is STATIC IP which I have configured.
The port number is set to 50000 on Server side and from the client side I connect to this IP on the port 50000.
I build and run the server side application and then try to connect to it by running a client side application.
One doubt about server side, server side application returns before I start the client side application, so should I keep it running (while(1);) so that I can connect to it from the client side?
What's going wrong am I forgetting something here? Help!
I'm pasting the very slightly modified (IP and port numbers are different) Beej's C code for Server side and Client side here:
Server.c
/*
** server.c
*/
#include <stdio.h>
#include <string.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netdb.h>
#include <arpa/inet.h>
#include <netinet/in.h>
int main(int argc, char *argv[])
{
// code for a server waiting for connections
// namely a stream socket on port 3490, on this host's IP
// either IPv4 or IPv6.
int sockfd;
struct addrinfo hints, *servinfo, *p;
int rv;
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_UNSPEC; // use AF_INET6 to force IPv6
hints.ai_socktype = SOCK_STREAM;
hints.ai_flags = AI_PASSIVE; // use my IP address
if ((rv = getaddrinfo(NULL, "50000", &hints, &servinfo)) != 0)
{
fprintf(stderr, "getaddrinfo: %s\n", gai_strerror(rv));
exit(1);
}
// loop through all the results and bind to the first we can
for(p = servinfo; p != NULL; p = p->ai_next)
{
if ((sockfd = socket(p->ai_family, p->ai_socktype,
p->ai_protocol)) == -1)
{
perror("socket");
continue;
}
if (bind(sockfd, p->ai_addr, p->ai_addrlen) == -1)
{
close(sockfd);
perror("bind");
continue;
}
break; // if we get here, we must have connected successfully
}
if (p == NULL)
{
// looped off the end of the list with no successful bind
fprintf(stderr, "failed to bind socket\n");
exit(2);
}
freeaddrinfo(servinfo); // all done with this structure
}
Client.c
/*
** client.c
*/
#include <stdio.h>
#include <string.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netdb.h>
#include <arpa/inet.h>
#include <netinet/in.h>
int main(int argc, char *argv[])
{
// code for a client connecting to a server
// namely a stream socket to www.example.com on port 80 (http)
// either IPv4 or IPv6
int sockfd;
struct addrinfo hints, *servinfo, *p;
int rv;
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_UNSPEC; // use AF_INET6 to force IPv6
hints.ai_socktype = SOCK_STREAM;
if ((rv = getaddrinfo("192.168.2.4", "50000", &hints, &servinfo)) != 0)
{
fprintf(stderr, "getaddrinfo: %s\n", gai_strerror(rv));
exit(1);
}
// loop through all the results and connect to the first we can
for(p = servinfo; p != NULL; p = p->ai_next)
{
if ((sockfd = socket(p->ai_family, p->ai_socktype,
p->ai_protocol)) == -1)
{
perror("socket");
continue;
}
if (connect(sockfd, p->ai_addr, p->ai_addrlen) == -1)
{
close(sockfd);
perror("connect");
continue;
}
break; // if we get here, we must have connected successfully
}
if (p == NULL)
{
// looped off the end of the list with no connection
fprintf(stderr, "failed to connect\n");
exit(2);
}
freeaddrinfo(servinfo); // all done with this structure
}
Your server code is missing listen() and accept() code to "wait" for a connection by calling listen() and then performing an accept() to accept new connections. Doesn't the example you are using show how to do that? Typically you will also fork a new thread for each new connection.
See http://www.linuxhowtos.org/C_C++/socket.htm for more information.
Here's a link to a more complete implementation: http://www.thegeekstuff.com/2011/12/c-socket-programming/
Yes, you need to keep the server program running. In your server program you have created the socket using socket() and bound to an address bind(), now you need to start listening for incoming connections. This is done with the listen() call. Once the socket is listening for incoming connections you have to use the accept() call to actually accept a connection and get the socket for communication with that particular client.
As a quick example, after the freeaddrinfo you could add the following code:
listen(sockfd, 8); /* allow up to 8 pending connections, i.e. created but not accepted */
while (1) {
int clientfd;
struct sockaddr_in client_addr;
socklen_t client_addr_len = sizeof(struct sockaddr_in);
clientfd = accept(sockfd, &client_addr, &client_addr_len);
/* communicate with client by send()/recv() or write()/read() on clientfd */
close(clientfd);
}
This code has the deficiency that only one client at a time is handled. There are a few ways to handle multiple simultaneous clients: multiple processes using fork(), multiple threads, or polling. Each of these approaches are, in my opinion, outside the scope of this question.
Please look at your Server.c file: it does not call listen() at all!
If the targeted server does not listen on the specified port, then it would return RST packet upon receiving SYN packet from the client, so connect() would return with "Connection refused".
Normal flow of functions at server side is socket -> bind -> listen -> accept :
getaddrinfo();
socket();
bind();
listen();
/* accept() goes here */
Please refer to https://beej.us/guide/bgnet/html/multi/syscalls.html#listen
I was getting the error "connection refused" due to not having the entry of remote host in /etc/hosts file. Entry should be present both side. In Client machines /etc/hosts, there should be a entry of server machine and vice versa in below pattern.
<ip address> <hostname with domain> <alias hostname>
This solved the error which i was in getaddrinfo() function.
Related
I have a C socket that listens on port 1001 on localhost. I also have the client code that connects to port 1001 on the ip 127.0.0.1. If I send the client's code to my friend, how could he have access to my machine when we would be on different networks? Is it possible for me just by changing the server code to make my public IP open for connections on port 1001? Below is the simple server code:
obs: I learning C.
#include <stdio.h>
#include <string.h>
#include <sys/socket.h>
#include <arpa/inet.h>
#include <unistd.h>
#define BUFSIZE 256
short SocketCreate(void)
{
short hSocket;
printf("Create the socket\n");
hSocket = socket(AF_INET, SOCK_STREAM, 0);
// close(hSocket);
return hSocket;
}
int BindCreatedSocket(int hSocket)
{
int iRetval = -1, ClientPort = 1001;
struct sockaddr_in remote = {0};
remote.sin_family = AF_INET;
remote.sin_addr.s_addr = htonl(INADDR_ANY);
remote.sin_port = htons(ClientPort);
iRetval = bind(hSocket, (struct sockaddr *)&remote, sizeof(remote));
return iRetval;
}
int main(int argc, char *argv[])
{
int socket_desc, sock, clientLen;
struct sockaddr_in client;
char client_message[200] = {0}, message[9999] = {0};
char buf[BUFSIZE];
socket_desc = SocketCreate();
if (socket_desc == -1)
{
printf("Could not create socket");
return 1;
}
printf("Socket created\n");
if (BindCreatedSocket(socket_desc) < 0)
{
perror("bind failed.");
return 1;
}
printf("Waiting for incoming connections...\n");
listen(socket_desc, 3);
while (1)
{
clientLen = sizeof(struct sockaddr_in);
sock = accept(socket_desc, (struct sockaddr *)&client, (socklen_t *)&clientLen);
if (sock < 0)
{
perror("accept failed");
return 1;
}
// printf("Connection accepted\n");
memset(client_message, '\0', sizeof(client_message));
memset(message, '\0', sizeof(message));
if (recv(sock, client_message, 200, 0) < 0)
{
printf("recv failed");
break;
}
if (strcmp(client_message, "exitserver") == 0)
{
close(socket_desc);
close(sock);
break;
}
}
return 0;
}
I have a C socket that listens on port 1001 on localhost. I also have
the client code that connects to port 1001 on the ip 127.0.0.1. If I
send the client's code to my friend, how could he have access to my
machine when we would be on different networks?
They couldn't. Address 127.0.0.1 is a loopback address. Packets sent to that address are always directed to the machine that sent them.
Is it possible for me just by changing the server code to make my
public IP open for connections on port 1001?
Do you have a public IP? 127.0.0.1 certainly isn't one, and most people with consumer-grade internet service don't have one. If you did have one, you probably would have had to make special arrangements to get it, and you would probably be paying extra for the privilege.
But supposing that you did have a public IP or that you made arrangements to get one, no, you cannot ensure that a port is open via your server program. Your program can listen on that address without much fuss, but you have to consider also firewalls -- probably one on your local machine and one at your local router, at least.
Also, before you set up a public server, you would be wise to check your ISP's policy and user agreement. It is not uncommon for ISPs to forbid running outward-facing services on consumer internet connections. They typically want you to pay more for that privilege, and that also makes it easier for ISPs to police their networks.
Using BJ's talker.c code as a template:
http://beej.us/guide/bgnet/examples/talker.c
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <errno.h>
#include <string.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <netdb.h>
#define SERVERPORT "4950" // the port users will be connecting to
int main(int argc, char *argv[])
{
int sockfd;
struct addrinfo hints, *servinfo, *p;
int rv;
int numbytes;
struct sockaddr_storage their_addr;
socklen_t addr_len;
addr_len = sizeof their_addr;
if (argc != 3) {
fprintf(stderr,"usage: talker hostname message\n");
exit(1);
}
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_UNSPEC;
hints.ai_socktype = SOCK_DGRAM;
if ((rv = getaddrinfo(argv[1], SERVERPORT, &hints, &servinfo)) != 0) {
fprintf(stderr, "getaddrinfo: %s\n", gai_strerror(rv));
return 1;
}
// loop through all the results and make a socket
for(p = servinfo; p != NULL; p = p->ai_next) {
if ((sockfd = socket(p->ai_family, p->ai_socktype,
p->ai_protocol)) == -1) {
perror("talker: socket");
continue;
}
break;
}
if (p == NULL) {
fprintf(stderr, "talker: failed to create socket\n");
return 2;
}
if ((numbytes = sendto(sockfd, argv[2], strlen(argv[2]), 0,
p->ai_addr, p->ai_addrlen)) == -1) {
perror("talker: sendto");
exit(1);
}
freeaddrinfo(servinfo);
printf("talker: sent %d bytes to %s\n", numbytes, argv[1]);
//============== Added Code for recvfrom() (pseudocode-ish) =============
if ((numbytes = recvfrom(sockfd, buf, MAXBUFLEN , 0, (struct sockaddr *)&their_addr, &addr_len)) == -1)
{
close(sockfd);
perror("talker: recvfrom");
exit(1);
}
close(sockfd);
printf("Got packet\n");
//============== End Added Code for recvfrom() =============
close(sockfd);
return 0;
}
I have a requirement whereby the client UDP process that talks to the server must use a fixed, known source port number. In this case, assume it's SERVERPORT (4950). The server then responds to that port number. Yes, this is unusual as most servers respond to the ephemeral port number that the system assigns to the sender.
After sending a packet using sendto(), I listen for a response using recvfrom(). That's the (pseudo)code I added in the above example.
All my searches online point to using bind() but that code is usually on the server side. I haven't found a way to bind on the client side using the modern getaddrinfo() method. I tried to add a bind() right after the socket() setup but that wouldn't work because p is a server-side structure (derived from the hints structure that uses the server IP address) and I get a bind Error:
Error 99 (Cannot assign requested address)
code added:
bind(sockfd, p->ai_addr, p->ai_addrlen)
I want to do this in a way that will work for both IPv4 and IPv6.
I've seen other examples whereby a local/source sockaddr_in structure is filled out with the client's information and that is used in the bind, but those are IPv4 or IPv6 specific.
Can someone please show me how to properly update the talker.c code to sendto() and recvfrom() a UDP server using a fixed source port number? Assume that the server is immutable.
The server then responds to that port number. Yes, this is unusual
There is nothing unusual about that. This is how most UDP servers are meant to work. They always respond to the sender's port. They have no concept whether that port is fixed or ephemeral, that is for the sender to decide. Unless a particular protocol dictates that responses are to be sent to a different port, which is not common.
All my searches online point to using bind()
Correct, that is what you need in this situation.
but that code is usually on the server side.
There is nothing preventing a client from using bind().
I haven't found a way to bind on the client side using the modern getaddrinfo() method.
It is the exact same as on the server side, except that you have to bind to a specific IP address, you can't bind to 0.0.0.0 or ::0 like you can with a server socket.
I tried to add a bind() right after the socket() setup but that wouldn't work
Yes, it does. The problem is that you are using the SAME IP address for both binding and sending, and that will not work. You need to bind to the CLIENT's IP address and then send to the SERVER's IP address.
because p is a server-side structure (derived from the hints structure that uses the server IP address)
You are misusing p. You can't bind() a client socket to the server's IP address (you need to use connect() for that instead). You need to bind() a client socket to an IP address that is local to the client's machine. Just like you have to bind() a server socket to an IP address that is local to the server machine.
Remember, a socket is associated with a pair of IP addresses. bind() establishes the socket's LOCAL IP address. connect() establishes the socket's REMOTE IP address.
I want to do this in a way that will work for both IPv4 and IPv6.
You can't create a single client socket for both protocols. You need separate sockets for each protocol (on the server side, you can create a single socket for both protocols, if your platform supports dual-stack sockets).
I've seen other examples whereby a local/source sockaddr_in structure is filled out with the client's information and that is used in the bind, but those are IPv4 or IPv6 specific.
Yes, because you will be sending a packet using EITHER IPv4 OR IPv6, you can't send a packet using both protocols at the same time (a dual-stack socket can receive packets from either protocol, though).
Can someone please show me how to properly update the talker.c code to sendto() and recvfrom() a UDP server using a fixed source port number . Assume that the server is immutable
Try something like this:
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <errno.h>
#include <string.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <netdb.h>
#include <stdbool.h>
#define LOCALPORT "4950" // the port users will be sending from
#define SERVERPORT "4950" // the port users will be connecting to
#define MAXBUFLEN 65535
int main(int argc, char *argv[])
{
int sockfd;
struct addrinfo hints, *myinfo, *servinfo, *pserv, *plocal;
int rv;
int numbytes;
char buf[MAXBUFLEN];
char ipstr[INET6_ADDRSTRLEN];
fd_set readfds;
struct timeval tv;
bool stop = false;
if (argc < 3) {
fprintf(stderr, "usage: talker destaddr message [localaddr]\n");
return 1;
}
// get all of the server addresses
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_UNSPEC;
hints.ai_socktype = SOCK_DGRAM;
hints.ai_protocol = IPPROTO_UDP;
if ((rv = getaddrinfo(argv[1], SERVERPORT, &hints, &servinfo)) != 0) {
fprintf(stderr, "getaddrinfo: %s\n", gai_strerror(rv));
return 2;
}
// loop through all the server addresses
for(pserv = servinfo; (pserv != NULL) && (!stop); pserv = pserv->ai_next) {
memset(ipstr, 0, sizeof(ipstr));
switch (pserv->ai_family)
{
case AF_INET:
inet_ntop(AF_INET, &(((struct sockaddr_in*)pserv->ai_addr)->sin_addr), ipstr, INET_ADDRSTRLEN);
break;
case AF_INET6:
inet_ntop(AF_INET6, &(((struct sockaddr_in6*)pserv->ai_addr)->sin6_addr), ipstr, INET6_ADDRSTRLEN);
break;
}
printf("talker: trying to send message to %s\n", ipstr);
// get all of the matching local addresses
memset(&hints, 0, sizeof hints);
hints.ai_family = pserv->ai_family;
hints.ai_socktype = pserv->ai_socktype;
hints.ai_protocol = pserv->ai_protocol;
if ((rv = getaddrinfo(argc > 3 ? argv[3] : NULL, LOCALPORT, &hints, &myinfo)) != 0) {
fprintf(stderr, "getaddrinfo: %s\n", gai_strerror(rv));
continue;
}
// loop through all the local addresses, sending the
// message from each one until a reply is received
for(plocal = myinfo; (plocal != NULL) && (!stop); plocal = plocal->ai_next) {
if ((sockfd = socket(plocal->ai_family, plocal->ai_socktype, plocal->ai_protocol)) == -1) {
perror("socket");
continue;
}
memset(ipstr, 0, sizeof(ipstr));
switch (plocal->ai_family)
{
case AF_INET:
inet_ntop(AF_INET, &(((struct sockaddr_in*)plocal->ai_addr)->sin_addr), ipstr, INET_ADDRSTRLEN);
break;
case AF_INET6:
inet_ntop(AF_INET6, &(((struct sockaddr_in6*)plocal->ai_addr)->sin6_addr), ipstr, INET6_ADDRSTRLEN);
break;
}
printf("talker: binding to %s\n", ipstr);
if (bind(sockfd, plocal->ai_addr, plocal->ai_addrlen) == -1) {
perror("bind");
close(sockfd);
continue;
}
// make sure this server address is the only one we talk to
if (connect(sockfd, pserv->ai_addr, pserv->ai_addrlen) == -1) {
perror("connect");
close(sockfd);
continue;
}
if ((numbytes = send(sockfd, argv[2], strlen(argv[2]), 0)) == -1) {
perror("send");
close(sockfd);
continue;
}
printf("talker: sent %d bytes\n", numbytes);
FD_ZERO(&readfds);
FD_SET(sockfd, &readfds);
tv.tv_sec = 5;
tv.tv_usec = 0;
rv = select(sockfd+1, &readfds, NULL, NULL, &tv);
if (rv == -1)
{
perror("select");
close(sockfd);
continue;
}
if (rv == 0)
{
printf("talker: no reply for 5 seconds\n");
close(sockfd);
continue;
}
if ((numbytes = recv(sockfd, buf, MAXBUFLEN, 0)) == -1)
{
perror("recv");
close(sockfd);
continue;
}
printf("talker: received %d bytes\n", numbytes);
close(sockfd);
stop = true;
break;
}
freeaddrinfo(myinfo);
}
freeaddrinfo(servinfo);
close(sockfd);
if (!stop) {
fprintf(stderr, "talker: failed to communicate with server\n");
return 3;
}
return 0;
}
I am trying to understand how the backlog parameter in int listen(int sockfd, int backlog); affects how new connections are handled.
Here is my server program.
/* server.c */
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include <sys/socket.h>
#include <arpa/inet.h>
#include <netdb.h>
int main()
{
int sockfd;
int ret;
int yes = 1;
struct addrinfo hints, *ai;
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_INET;
hints.ai_socktype = SOCK_STREAM;
hints.ai_flags = AI_PASSIVE;
if ((ret = getaddrinfo(NULL, "8000", &hints, &ai)) == -1) {
fprintf(stderr, "getaddrinfo: %s\n", gai_strerror(ret));
return 1;
}
sockfd = socket(ai->ai_family, ai->ai_socktype, ai->ai_protocol);
if (sockfd == -1) {
perror("server: socket");
return 1;
}
if (setsockopt(sockfd, SOL_SOCKET, SO_REUSEADDR, &yes, sizeof yes) == -1) {
perror("server: setsockopt");
close(sockfd);
return 1;
}
if (bind(sockfd, ai->ai_addr, ai->ai_addrlen) == -1) {
perror("server: bind");
close(sockfd);
return 1;
}
freeaddrinfo(ai);
if (listen(sockfd, 2) == -1) {
perror("server: listen");
close(sockfd);
return 1;
}
printf("server: listening ...\n");
printf("server: sleep() to allow multiple clients to connect ...\n");
sleep(10);
printf("server: accepting ...\n");
while (1) {
int connfd;
struct sockaddr_storage client_addr;
socklen_t client_addrlen = sizeof client_addr;
char buffer[1024];
int bytes;
connfd = accept(sockfd, (struct sockaddr *) &client_addr, &client_addrlen);
if (connfd == -1) {
perror("server: accept");
continue;
}
if ((bytes = recv(connfd, buffer, sizeof buffer, 0)) == -1) {
perror("server: recv");
continue;
}
printf("server: recv: %.*s\n", (int) bytes, buffer);
close(connfd);
}
return 0;
}
Here is my client program.
/* client.c */
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include <sys/socket.h>
#include <arpa/inet.h>
#include <netdb.h>
int main(int argc, char **argv)
{
int sockfd;
int ret;
struct addrinfo hints, *ai;
if (argc != 2) {
fprintf(stderr, "usage: %s MSG\n", argv[0]);
return 1;
}
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_INET;
hints.ai_socktype = SOCK_STREAM;
if ((ret = getaddrinfo(NULL, "8000", &hints, &ai)) == -1) {
fprintf(stderr, "client: getaddrinfo: %s\n", gai_strerror(ret));
return 1;
}
sockfd = socket(ai->ai_family, ai->ai_socktype, ai->ai_protocol);
if (sockfd == -1) {
perror("client: socket");
return 1;
}
if (connect(sockfd, ai->ai_addr, ai->ai_addrlen) == -1) {
perror("client: connect");
close(sockfd);
return -1;
}
printf("client: connected\n");
if (send(sockfd, argv[1], strlen(argv[1]), 0) == -1) {
perror("client: send");
close(sockfd);
return -1;
}
printf("client: send: %s\n", argv[1]);
freeaddrinfo(ai);
close(sockfd);
return 0;
}
I compile and run these programs with the following script.
# run.sh
gcc -std=c99 -Wall -Wextra -Wpedantic -D_DEFAULT_SOURCE server.c -o server
gcc -std=c99 -Wall -Wextra -Wpedantic -D_DEFAULT_SOURCE client.c -o client
./server &
sleep 1
./client hello1 &
sleep 1
./client hello2 &
sleep 1
./client hello3 &
sleep 1
./client hello4 &
sleep 1
./client hello5 &
sleep 5
pkill server
When I run the above script, I get this output.
$ sh run.sh
server: listening ...
server: sleep() to allow multiple clients to connect ...
client: connected
client: send: hello1
client: connected
client: send: hello2
client: connected
client: send: hello3
client: connected
client: send: hello4
client: connected
client: send: hello5
server: accepting ...
server: recv: hello1
server: recv: hello2
server: recv: hello3
The output shows that while the server was sleeping between listen() and accept(), all five clients could successfully connect() and send() to the server. However, the server could accept() and recv() three clients only.
I don't understand the following.
The server program invokes listen() with the backlog parameter as 2. Why did all five clients succeed in connect()-ing then? I was expecting only 2 connect()s to be successful.
Why was the server able to accept() and recv() from 3 clients instead of 2?
The server program invokes listen() with the backlog parameter as 2.
Why did all five clients succeed in connect()-ing then?
backlog parameter is only a hint for listen(). From POSIX doc:
The backlog argument provides a hint to the implementation which the
implementation shall use to limit the number of outstanding
connections in the socket's listen queue. Implementations may impose a
limit on backlog and silently reduce the specified value. Normally, a
larger backlog argument value shall result in a larger or equal length
of the listen queue. Implementations shall support values of backlog
up to SOMAXCONN, defined in .
When a client connects to the listening port, depending on the implementation of the socket stack, it may either:
hold the pending connection in the backlog, and complete the 3-way TCP handshake only when accept() is called to remove that client from the backlog. This is the behavior you are expecting, and is how older systems behaved.
complete the handshake right away in the background and then store the fully connected connection in the backlog until accept() removes it. This is the behavior your example appears to be exhibiting, and is not uncommon in modern systems.
According to the Linux manpage for listen():
The behavior of the backlog argument on TCP sockets changed with Linux 2.2. Now it specifies the queue length for completely established sockets waiting to be accepted, instead of the number of incomplete connection requests. The maximum length of the queue for incomplete sockets can be set using /proc/sys/net/ipv4/tcp_max_syn_backlog. When syncookies are enabled there is no logical maximum length and this setting is ignored. See tcp(7) for more information.
So, in your case, all 5 connections are likely being completed in the background before you start calling accept(), thus allowing the clients to call send() (and may do so before they can detect that some of the connections are being dropped), but not all of the connections are able to remain in the backlog due to its small size.
The issue actually seems to be that the test isn't isolating the backlog that it assumes to be testing.
The test code in the question seems to use "blocking" sockets and concurrency is invoked by demonizing the client test, which might explain how another client "got in".
To correctly test the issue, it's important to have a concurrent model where we know how much stress is exerted on the system at any point in time.
It's also important that we only clear the backlog once, without waiting for the kernel to refill the backlog we allocated with the kernel's layer backlog.
Attached is a concurrent (threaded) client+server that both listens, connects (to itself) and prints out the messages.
This design makes it clear how much stress (5 connections) the server experiences concurrently.
To make it a bit clearer, I chose to avoid "blocking" sockets as far as the server thread is concerned. This way we can accept everything in the backlog and get a notification (an error value) when the backlog is empty.
On my platform (macOS), the results show that only two clients manage to connect to the server, conforming to the listen(socked, 2) backlog specification.
All the other clients fail because the kernel drops the connection when it can't push it into the (full) backlog... though we don't know the connections were dropped until read is attempted... also some of my error checks aren't perfect):
server: listening ...
server: sleep() to allow multiple clients to connect ...
client: connected
client: connected
client: connected
client: connected
client: connected
client: read error: Connection reset by peer
client: read error: Connection reset by peer
client: read error: Connection reset by peer
server: accepting ...
client 3: Hello World!
client 5: Hello World!
The connected clients (3 & 5 in this example) are dependent on the thread scheduler, so every time the test is performed a different pair of clients will manage to connect.
It's true that connect returns successfully, but connect seems to be optimistically implemented by the accepting kernel, as pointed out in #RemyLebeau's answer. On some systems (i.e. Linux and macOS), the kernel will complete the TCP/IP handshake before attempting to attach the connection to our listening socket's backlog (OR dropping it if the backlog is full).
This is easy to see on my system's output, where the "server: accepting..." message arrives after both the "connect" confirmation and the "Connection reset by peer" events.
The code for the test was:
#include <limits.h>
#include <pthread.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <arpa/inet.h>
#include <fcntl.h>
#include <netdb.h>
#include <sys/socket.h>
void *server_threard(void *arg);
void *client_thread(void *arg);
int main(void) {
/* code */
pthread_t threads[6];
if (pthread_create(threads, NULL, server_threard, NULL))
perror("couldn't initiate server thread"), exit(-1);
sleep(1);
for (size_t i = 1; i < 6; i++) {
if (pthread_create(threads + i, NULL, client_thread, (void *)i))
perror("couldn't initiate client thread"), exit(-1);
}
for (size_t i = 0; i < 6; i++) {
pthread_join(threads[i], NULL);
}
return 0;
}
/* will start listenning, sleep for 5 seconds, then accept all the backlog and
* finish */
void *server_threard(void *arg) {
(void)(arg);
int sockfd;
int ret;
struct addrinfo hints, *ai;
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_INET;
hints.ai_socktype = SOCK_STREAM;
hints.ai_flags = AI_PASSIVE;
if ((ret = getaddrinfo(NULL, "8000", &hints, &ai)) == -1) {
fprintf(stderr, "getaddrinfo: %s\n", gai_strerror(ret));
exit(1);
}
sockfd = socket(ai->ai_family, ai->ai_socktype, ai->ai_protocol);
if (sockfd == -1) {
perror("server: socket");
exit(1);
}
ret = 1;
if (setsockopt(sockfd, SOL_SOCKET, SO_REUSEADDR, &ret, sizeof ret) == -1) {
perror("server: setsockopt");
close(sockfd);
exit(1);
}
if (bind(sockfd, ai->ai_addr, ai->ai_addrlen) == -1) {
perror("server: bind");
close(sockfd);
exit(1);
}
freeaddrinfo(ai);
/* Set the server to non_blocking state */
{
int flags;
if (-1 == (flags = fcntl(sockfd, F_GETFL, 0)))
flags = 0;
// printf("flags initial value was %d\n", flags);
if (fcntl(sockfd, F_SETFL, flags | O_NONBLOCK) < 0) {
perror("server: to non-block");
close(sockfd);
exit(1);
}
}
if (listen(sockfd, 2) == -1) {
perror("server: listen");
close(sockfd);
exit(1);
}
printf("server: listening ...\n");
printf("server: sleep() to allow multiple clients to connect ...\n");
sleep(5);
printf("server: accepting ...\n");
int connfd;
struct sockaddr_storage client_addr;
socklen_t client_addrlen = sizeof client_addr;
/* accept up all connections. we're non-blocking, -1 == no more connections */
while ((connfd = accept(sockfd, (struct sockaddr *)&client_addr,
&client_addrlen)) >= 0) {
if (write(connfd, "Hello World!", 12) < 12)
perror("server write failed");
close(connfd);
}
close(sockfd);
return NULL;
}
void *client_thread(void *arg) {
(void)(arg);
int sockfd;
int ret;
struct addrinfo hints, *ai;
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_INET;
hints.ai_socktype = SOCK_STREAM;
if ((ret = getaddrinfo(NULL, "8000", &hints, &ai)) == -1) {
fprintf(stderr, "client: getaddrinfo: %s\n", gai_strerror(ret));
exit(1);
}
sockfd = socket(ai->ai_family, ai->ai_socktype, ai->ai_protocol);
if (sockfd == -1) {
perror("client: socket");
exit(1);
}
if (connect(sockfd, ai->ai_addr, ai->ai_addrlen) < 0) {
perror("client: connect error");
close(sockfd);
fprintf(stderr, "client number %lu FAILED\n", (size_t)arg);
return NULL;
}
printf("client: connected\n");
char buffer[128];
if (read(sockfd, buffer, 12) < 12) {
perror("client: read error");
close(sockfd);
} else {
buffer[12] = 0;
fprintf(stderr, "client %lu: %s\n", (size_t)arg, buffer);
}
return NULL;
}
regarding:
listen( sock, 2).
and its' ability to handle 3 connections.
The 2 is how many connections can be in the queue.
When listen() first returns, there are the current connection and room in the queue for 2 more.
I.E. at total of 3 connections.
There are two queues for backlog, one is for connections which have finished three-way handshake and the other is for TCP in SYN_RCVD state which havan't received ACK from the remote client. Sum of these two queue size must be less equal than backlog. When you call accept, os retrieve one ESTABLISHED connection from connected queue. So you can accept too many connections from the established queue. This is not inconsistent with the backlog.
As your code sleep one second between each client, clients have time to finish and close their connections before the next one comes.
So the queue on server side (this is what backlog argument controls) is always empty.
Try again without "sleep" statements.
I wrote a simple application which connects to a given server on a given port. When the port is open, everything is ok, I got the message about the established connection. However, when the port is closed, nothing happens, my program does not show me the information about it.
I test my program using my remote server accessible via the Internet. How can I improve this?
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <unistd.h>
int main(int argc, char **argv) {
char * ip_addr;
int port;
int sockfd;
struct sockaddr_in server_addr;
if (argc != 3)
{
fprintf(stderr,"Usage: ./canconnect ip port\n");
exit(1);
}
ip_addr = argv[1];
port = atoi(argv[2]);
if (port <= 0)
{
fprintf(stderr,"error: invalid port\n");
exit(1);
}
if ((sockfd = socket(AF_INET, SOCK_STREAM, 0)) == -1)
{
perror("socket");
exit(1);
}
server_addr.sin_family = AF_INET;
server_addr.sin_port = htons(port);
inet_aton(ip_addr, &server_addr.sin_addr);
if (connect(sockfd, (struct sockaddr *)&server_addr, sizeof(struct sockaddr)) == -1)
{
perror("connect");
printf("Port %d is closed on server %s.\n", port, ip_addr);
exit(1);
}
else
{
printf("Connection established. Port %d is open on server %s.\n", port, ip_addr);
}
close(sockfd);
return 0; }
Given that the program actually prints some data after some time, it's most certainly has to do with timeout.
In order to finally decide that the host or the port is unreadable or the connection couldn't be established for any other reason, connect performs several attempts to connect and returns an error after a certain amount of time - the timeout.
The value of timeout can be changed to any value you want using setsockopt:
struct timeval timeout;
timeout.tv_sec = 3; // wait for three seconds
timeout.tv_usec = 0;
// set up receive timeout
if (setsockopt(sockfd, SOL_SOCKET, SO_RCVTIMEO, (char *)&timeout,
sizeof(timeout)) < 0)
printf("[!] setsockopt failed\n");
If the remote server is reachable, but neither accepts nor refuses the connection, then connect() will block until the network connection attempt times out. As the Linux manual page for connect(2) puts it:
Note that for IP sockets the timeout may be very long when syncookies are enabled on the server.
Indeed, it is a well-known defense against port scans to attempt to elicit that behavior intentionally. You might be able to get quicker failures by using setsockopt() to set a receive timeout, but the docs are inconsistent on whether that will work for connect().
Is it possible that a TCP server program can listen on two different socket interface?
Problem Statement:
I've a problem statement where the TCP server will be having two interfaces:
Interface I: For accepting generic data from TCP client (IP address 192.168.5.10:2000)
Interface II: Management Interface for the server (IP address 192.168.5.11:2000)
Interface I: This interface shall receive data from TCP client, processes them & send it back to client.
Interface II: This interface shall receive commands (meant for Servers management purpose). This commands most probably would be sent through telnet.
Current Status:
I already have a thread based TCP server program where I've "Interface I" up & running(I'm able to receive data from multiple clients, process them & send it back)
Can anyone give me some pointers/prototype example on how to add "Interface II" to my TCP server program?
NOTE: TCP server program is written in C programming language for Linux OS
UPDATE
Below is the code fragment I've written so far for listening on one socket. I tried modifying it for listening over two sockets as you've directed but I'm facing trouble while trying to spawn a different thread for the other socket interface. Will it possible for you to modify this to listen on two sockets? It would be really helpful.
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <errno.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <pthread.h>
void *data_processing_thread(void *arg);
int main(int argc, char **argv)
{
int fdmax, listener, newfd, res;
int optval=1;
socklen_t addrlen;
int server_port = 4000;
/* master, temp file descriptor list */
fd_set *master, *read_fds;
/* client, server address */
struct sockaddr_in server_addr, client_addr;
pthread_t thread;
master = malloc(sizeof(fd_set));
read_fds = malloc(sizeof(fd_set));
FD_ZERO(master);
FD_ZERO(read_fds);
/* create endpoint for communication */
if ((listener = socket(AF_INET, SOCK_STREAM, 0)) == -1) {
perror("failed to create listener\n");
return -1;
}
/* check if address is already in use? */
if (setsockopt(listener, SOL_SOCKET, SO_REUSEADDR, &optval,
sizeof(int)) == -1) {
perror("socket address already in use!\n");
return -1;
}
/* bind */
server_addr.sin_family = AF_INET;
server_addr.sin_addr.s_addr = INADDR_ANY;
server_addr.sin_port = htons(server_port);
memset(&(server_addr.sin_zero), '\0', 8);
if (bind(listener, (struct sockaddr*)&server_addr,
sizeof(server_addr)) == -1) {
perror("failed to do the bind\n");
return -1;
}
/* listen for connect on sockets */
if (listen(listener, 10) == -1) {
perror("failed to listen on socket\n");
return -1;
}
/* add the listener to the master set */
FD_SET(listener, master);
/* keep track of biggest file descriptor */
fdmax = listener;
while (1) {
read_fds = master;
/* wait till socket descriptor is ready for the operation */
if (select(fdmax+1, read_fds, NULL, NULL, NULL) == -1) {
perror("failed to do select() on socket\n");
return -1;
}
/* Run through existing data connections looking for data to be
* read */
int cnt;
int *accept_fd = 0;
for (cnt=0; cnt<=fdmax; cnt++) {
if (cnt == listener) {
if (FD_ISSET(cnt, read_fds)) {
addrlen = sizeof(client_addr);
if ((newfd = accept(listener, (struct sockaddr*)&client_addr, &addrlen)) == -1) {
perror("failed to accept incoming connection\n");
} else {
fprintf(stdout, "Server: Connection from client [%s] on socket [%d]\n",
inet_ntoa(client_addr.sin_addr), newfd);
accept_fd = malloc(sizeof(int));
*accept_fd = newfd;
if ((res = pthread_create(&thread, NULL, data_processing_thread, (void*)accept_fd)) != 0) {
perror("Thread creation failed\n");
free(accept_fd);
}
}
}
continue;
}
}
}
return 1;
}
void *data_processing_thread(void *arg)
{
int nbytes;
int *recv_fd = (int*)arg;
char *buffer = malloc(sizeof(char)*256);
while(1) {
fprintf(stdout, "Server: Waiting for data from socket fd %d\n", *recv_fd);
/* receive incoming data from comm client */
if ((nbytes = recv(*recv_fd, buffer, sizeof(buffer), 0)) <= 0) {
if (nbytes != 0) {
perror("failed to receive\n");
}
break;
} else {
fprintf(stdout, "Data received: %s\n", buffer);
}
}
close(*recv_fd);
free(recv_fd);
pthread_exit(0);
}
Create two listening sockets using socket().
Bind both to respective address/port using bind().
Make both listen using listen().
Add both listening sockets to a properly initialised fd_set typed variable using FD_SET().
Pass the fd_set to a call to select()
Upon select()'s return check the reason and perform the appropriate action, typically
either calling accept() on one of the both listening sockets and add the accepted socket (as returned by accept()) to the fd_set,
or if it's an accepted socket that had triggered select() to return, then call read(), write() or close() on it. If close()ing the socket also remove it from the fd_set using FD_CLR().
Start over with step 5.
Important note: The steps above are a rough scheme, not mentioning all possible all traps, so it is absolutly necessary to also read the related man-pages for each step carefully, to understand what is happening.
you can bind 0.0.0.0 which means binding all interfaces.
you can't bind two interfaces using only one socket.
you should create a new socket, and bind ti to interface II.