Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have written an network discovery code to discover the devices on the network.The code should discover the devices on the network even if the icmp echo response is disabled . I read about tcp ping from nmap where it sends TCP SYN packet and creates an half open connection to the remote device to check whether it is discoverable .In the code i made use of connect function from the windows socket library to check whether the host is discover able or not by establishing an socket connection on port 80 .

I disabled the ping on one of the test machines which runs windows 7 by following the link to check whether the machine is discover able or not but surprisingly the socket connection too failed with connection refused error . However nmap could discover the device through the TCP ping option (which is through default port 80 ) .

So my questions:

  1. Is what is the exact difference between both of them ?

  2. What are other alternate ways to discover the devices on the network.

This is just a sample code to establish an socket connection .

  #ifndef UNICODE
  #define UNICODE
  #endif

  #define WIN32_LEAN_AND_MEAN

  #include <winsock2.h>
  #include <ws2tcpip.h>
  #include <stdio.h>

 // Need to link with Ws2_32.lib
 #pragma comment(lib, "ws2_32.lib")

 int wmain()
 {

   // Initialize Winsock
    WSADATA wsaData;
    int i=0;
    char ip[20];
    int iResult = WSAStartup(MAKEWORD(2, 2), &wsaData);
    if (iResult != NO_ERROR) {
    wprintf(L"WSAStartup function failed with error: %d\n", iResult);
    return 1;
    }

    SOCKET ConnectSocket;
    sockaddr_in clientService;

do{

     // Create a SOCKET for connecting to server

      ConnectSocket = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
      if (ConnectSocket == INVALID_SOCKET) {
       wprintf(L"socket function failed with error: %ld\n", WSAGetLastError());
       WSACleanup();
       return 1;
     }

     // The sockaddr_in structure specifies the address family,
      // IP address, and port of the server to be connected to.

     printf("\n Enter the Ip Address : ");
     scanf("%s",ip);
     clientService.sin_family = AF_INET;
     clientService.sin_addr.s_addr = inet_addr(ip);
     clientService.sin_port = htons(80);

     // Connect to server

     iResult = connect(ConnectSocket, (SOCKADDR *) & clientService, sizeof (clientService));

    if (iResult == SOCKET_ERROR) {
    wprintf(L"connect function failed with  machine number %d  with error: %ld\n",i, WSAGetLastError());

    iResult = closesocket(ConnectSocket);
    if (iResult == SOCKET_ERROR)
        wprintf(L"closesocket function failed with  machine %d error: %ld\n",i, WSAGetLastError());
        return 1;
    }
    wprintf(L"\n Connected to  machine  %d : ",i);

    iResult = closesocket(ConnectSocket);
    if (iResult == SOCKET_ERROR) {
    wprintf(L"closesocket function failed with  machine %d error: %ld\n",i, WSAGetLastError());
    WSACleanup();
    return 1;
   }
   i++;

}while(i<2);
   WSACleanup(); 
   return 0;
 } 
share|improve this question
    
Could you provide your code? –  cuonglm Aug 7 '13 at 9:48
    
I have posted the code . –  Santhosh Pai Aug 7 '13 at 10:00

1 Answer 1

up vote 3 down vote accepted

Is what is the exact difference between both of them

Between what sorry? ICMP ping and TCP ping? - Well one uses the ICMP echo and reply messages which I believe you already know, the other tries to play on the negotiation of a 3-way TCP handshake.

If you mean what is the difference between TCP Ping in your code and nmap, you'd have to read the nmap code for that. Or another idea would be to run something like http://www.wireshark.org/to see what you application sends down the wire, and what nmap is sending down the wire, and comparing the results.

What are other alternate ways to discover the devices on the network.

Look through all the options in nmap and read some nmap tutorials, there are a few different discovery options there (research about how they work). One good method is trying to connect to TCP port 0 on machines. Different machines react in different ways, as this is an unusable socket (reserved). This is often used of OS identification (some OSs drop the packet, others send a RST back etc, so you can guess what OS might be assigned to that IP).

No method is fool proof though for host discovery. You can configure firewalls to react in all sorts of ways (silently drop packets, respond with a RST packet, send back ICMP prohibited packets, and so on). So factoring this in, host discovery can be very difficult when firewalls are involved. If you are on the same LAN as a host you can play around with arpping, RARPs etc, gratuitous ARPs, to try and get the MAC of the host, and possibly the IP. This can work even if they don't respond to pings etc. You can also play with sending IP packets to a broadcast address (255.255.255.255) and multicast addresses (this includes layer 2 traffic FF:FF:FF:FF:FF:FF etc).

It's a long subject, that's not super easy, so best of luck!

UPDATE - difference between TCP connection and TCP ping

The "TCP ping" here is just sending a TCP SYN packet to the destination host. It just runs the first part of the TCP 3 way handshake, you can see this on the Wikipedia page for TCP here. A TCP connection can be filtered by a firewall, i.e. It might see a SYN packet come through, a SYN-ACK returned, then before the final ACK is passed it filters it, to stop a TCP connection reaching the CONNECTED state. It is tracking the state of TCP connections. If you send a SYN packet on it's own, a firewall might not recognise this.

TCP SYN scanning, and various other scanning techniques are found on Wikipedia page here. The Wiki page describes it more succinctly than I can, so I shall quote that directly:

SYN scan is another form of TCP scanning. Rather than use the operating system's network functions, the port scanner generates raw IP packets itself, and monitors for responses. This scan type is also known as "half-open scanning", because it never actually opens a full TCP connection. The port scanner generates a SYN packet. If the target port is open, it will respond with a SYN-ACK packet. The scanner host responds with a RST packet, closing the connection before the handshake is completed. If the port is closed but unfiltered, the target will instantly respond with a RST packet.

share|improve this answer
    
Hi Javano , i wanted to know the difference between TCP connection and TCP ping . –  Santhosh Pai Aug 7 '13 at 10:34
    
@SanthoshPai Thanks for clarifying that, I have updated my answer. –  jwbensley Aug 7 '13 at 10:48
    
Thanks for the response .Its a very good answer –  Santhosh Pai Aug 7 '13 at 10:55
    
I have one basic question if the host ip adress is invalid or if an host does not exist do we get an RST reply ? . –  Santhosh Pai Aug 7 '13 at 11:40
1  
@SanthoshPai You should ask new quested in a new post. The answer though is no. A RST packet has to be send from a host, so if the host doesn't exist, it can't send a RST :) Also, if the IP is invalid, no host would receive a SYN to then reply with a RST packet. In the case of a firewall between you and the host, a firewall can be configured to intercept and respond with a RST on behalf of the host, to tell you to "go away". –  jwbensley Aug 7 '13 at 11:42

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.