在访问Web应用程序的服务器上获取客户端IP和主机名

Question

I have the following method but on some system it gives me all IP addresses ( IPv6 , IPv4 , and virtual IPv6 and virtual IPv4 ) and on a few system it gives me only IPv6 and virtual IPv6 but I need all IP's of client who hit the web-app to store in database.

Is there a problem with the system configuration?

public void Main()
{
    Username = System.Security.Principal.WindowsIdentity.GetCurrent().Name.ToString();

    computerName = System.Net.Dns.GetHostName(); 
    IPHostEntry ipEntry = System.Net.Dns.GetHostEntry(computerName); 

    IPAddress[] ipAddress = ipEntry.AddressList; 
    IPv6 = ipAddress[0].ToString(); 

    IPv4 = ipAddress[2].ToString(); 
    VirtualIPv4 = ipAddress[3].ToString(); 

    IFormatProvider theCultureInfo = new System.Globalization.CultureInfo("en-GB",true); 

    dt1 = Convert.ToDateTime(DateTime.Now,theCultureInfo).ToString("MM-dd-yyyy hh:mm tt");
}

Answer 1

Conceptual points:

1. Both servers and clients have IP addresses, that's how data gets sent (via IP datagrams)
2. Clients connect to a server using a single IP address (whether v4 or v6)... server binds to one or more IP addresses*

If you are still confused by this, see the bottom section

So, since you are getting a list of IP addresses, you must be doing something wrong. In this case, you are calling on the server a method to get the current IP address(es) of the server.

Username = System.Security.Principal.WindowsIdentity.GetCurrent().Name.ToString();

The above is where you go wrong - everything after that is pointless. You are running that code on the server, and WindowsIdentity.GetCurrent() is whoever is running the application.

Getting the client IP address

You want to get the Request IP address, which varies from framework to framework , but is something like HttpRequest.UserHostAddress

In ASP.NET, you can use:

1. HttpContext.Current.Request.UserHostAddress
2. HttpContext.Current.Request.ServerVariables["REMOTE_ADDR"]

What about IPv4 vs v6? Not your concern - the client uses one or the other, and so whatever it is, it will be in that value.

How to think about IP addressing wrt web apps

You have two players in this world: a client, who wants to get to the web app, and a web app, who serves clients.

The way a client identifies himself and the way a server identifies himself is via some address, usually an IP address , unless you are using an exotic networking stack.

Now, there are two versions of IP that are common: IPv4, the more common and older version, which looks like 123.45.67.89 and the other one, which is hexadecimal and ugly.

The address version corresponds to the IP version, the protocol version. To communicate with me via IPv6, we need IPv6 addresses. Likewise, to use IPv4, we need IPv4 addresses.

So, to start, we need a server IP or IPs to connect to. If multiple IPs are available for different versions, we pick the highest version that is available on both ends, so if the server has v4 only, the client must communicate via v4 only. If the client has both and the server is listening only on v6, then they will communicate via IPv6, and thus the client will use the IPv6 address to identify himself.

Finally, on both ends, keep in mind that IP addresses may not be the final word in identifying who is behind each end. That is because of the potential (and often reality) of layers in between the nodes that technically violate the "end to end principle," but in practice are very very common. In these scenarios, an IP address may be shared by multiple private computers (most common), a set of IP addresses may be shared by a set of computers (less common, used for servers in round-robin configuration), a firewall in between may only permit traffic one direction or at certain times of day, or forward based on the sender IP, etc.

Some examples

• Port forwarding: You have an Internet connection using a cable modem that is connected to a firewall/router/wifi device. Six different computers and phones connect to the firewall/router to share that one connection. They all share one IP, but traffic is routed back to them based on a port number assignment that is temporarily made by the firewall.
• NAT: This is a more general term for port forwarding and covers more exotic scenarios, but basically is a similar concept - you have one IP address (ingress) coming in, a firewall rewrites it (translation) so when you or the server or whoever gets it, it is translated to a routable address. On the way out (egress), it needs to be NAT'ted again, this time going from the IP addresses you understand to the IP addresses publicly-understood. (Outbound NAT)
• Round robin: M computers share N IP addresses; something in between rotates who gets what IP.