.NET Forum / Languages / C# / June 2007
Socket locking up after inital use
|
|
Thread rating:  |
Pain and headache - 29 Jun 2007 21:22 GMT Hi,
I am trying to get a webpage using a TcpSocket instead of a standard Webrequest. Initial, it works fine but after the 2 or 3 request the tcpclient I start to get the following error:
A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond
Here is the code making the call, this is the third iteration after trying several others with out sucess:
string contents = "GET " + pathQuery + " HTTP/1.1\r\n" + "Host: " + host + ":" + port + "\r\n" + "User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.8.1.4) Gecko/20070515 Firefox/2.0.0.4\r\n" + "Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5\r\n" + "Accept-Language: en-us,en;q=0.5\r\n" + "Accept-Encoding: gzip,deflate\r\n" + "Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7\r\n" + "Keep-Alive: 300\r\n" + "Connection: keep-alive\r\n\r\n";
Byte[] requestObject = Encoding.ASCII.GetBytes(contents.ToCharArray()); try { stream = socket.GetStream();
if (stream.CanWrite) { stream.Write(requestObject, 0, contents.Length); }
MemoryStream memoryStream = new MemoryStream();
if (stream.CanRead) { int data = stream.ReadByte(); while (data != -1) { memoryStream.WriteByte((byte) data); data = stream.ReadByte(); } results = Encoding.ASCII.GetString(memoryStream.ToArray()); }
} catch (Exception e) { Console.WriteLine(Encoding.ASCII.GetString(memoryStream.ToArray())); throw e; } finally { stream.Close(); socket.Close(); }
Peter Duniho - 29 Jun 2007 21:47 GMT > I am trying to get a webpage using a TcpSocket instead of a standard > Webrequest. Initial, it works fine but after the 2 or 3 request the [quoted text clipped - 3 lines] > respond after a period of time, or established connection failed because > connected host has failed to respond [...] Some thoughts:
* You say you get an error trying to make the connection. But you didn't bother to post any code related to creating or connecting your TCP socket.
* As far as I know, there's no such thing as a TcpSocket, nor is there a Socket.GetStream() method. You can create a Socket instance that encapsulates a TCP socket, and you can create a NetworkStream instance using a Socket instance. Is that what you did? If so, why does your post and code say something else? And if not, what _did_ you do?
All that said, assuming your code normally works, you may be running into some kind of anti-DoS defense, depending on how you're using the code. If you are repeatedly trying to make the same request to the same HTTP server, it may detect that case and stop responding, at least for some time.
If you can post a concise-but-complete example of code, client _and_ server, that reliably reproduces the problem, someone here may be able to provide better advice. Absent that, there's not much in your post to go on, and in fact what you posted doesn't make much sense in the context of regular .NET/C# programming (since you mention a class and a method that are simply not present in the basic .NET Framework).
Pete
Pain and headache - 29 Jun 2007 22:02 GMT I am very sorry for the confusing. I was trying to be brief. Here is the class, and I renamed the socket to client for easyier reading. I am using TCPTrace at the moment to test this, and it gets and sends the proper responses so I do not believe it is a Dos prevention of some kind.
using System; using System.IO; using System.Net.Sockets; using System.Text;
namespace ReanneCorp.Library.V2.Common.WebUtils { public class PullWebsiteClass { private Uri websiteUri; private TcpClient client = null; private string results = ""; private int port; private string host; private string pathQuery;
public Uri WebsiteUri { get { return websiteUri; } set { websiteUri = value; } }
public TcpClient Client { get { return client; } set { client = value; } }
public NetworkStream Stream { get { return stream; } set { stream = value; } }
public string Results { get { return results; } set { results = value; } }
public PullWebsiteClass(Uri websiteUri) { this.websiteUri = websiteUri; InitConnection(); } private void InitConnection() { port = websiteUri.Port; host = websiteUri.Host; pathQuery = websiteUri.PathAndQuery; client = new TcpClient(); client.ReceiveTimeout = 10000; try { client.Connect(host, port); } catch(SocketException ex) { throw ex; } } public void MakePageRequest() { string contents = "GET " + pathQuery + " HTTP/1.1\r\n" + "Host: " + host + ":" + port + "\r\n" + "User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.8.1.4) Gecko/20070515 Firefox/2.0.0.4\r\n" + "Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5\r\n" + "Accept-Language: en-us,en;q=0.5\r\n" + "Accept-Encoding: gzip,deflate\r\n" + "Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7\r\n" + "Keep-Alive: 300\r\n" + "Connection: keep-alive\r\n\r\n"; MemoryStream memoryStream = new MemoryStream();
Byte[] requestObject = Encoding.ASCII.GetBytes(contents.ToCharArray()); try { stream = client.GetStream();
if (stream.CanWrite) { stream.Write(requestObject, 0, contents.Length); } if (stream.CanRead) { int data = stream.ReadByte(); while (data != -1) { memoryStream.WriteByte((byte) data); data = stream.ReadByte(); } results = Encoding.ASCII.GetString(memoryStream.ToArray()); }
if (results.Contains("HTTP/1.1 302 Moved Temporarily")) { client.Close(); string redirect = results.Remove(0, results.IndexOf("Location: ") + 10);
redirect = redirect.Substring(0, redirect.IndexOf("Content-Length:") - 1).Trim();
PullWebsiteClass newRequest = new PullWebsiteClass(new Uri(redirect)); newRequest.MakePageRequest(); results = newRequest.Results; } } catch (Exception e) { throw e } finally { memoryStream.Close(); client.Close(); } } } }
Here is the calling code:
[Test] public void CheckHeaders() { PullWebsiteClass getPage = new PullWebsiteClass(new Uri("http://localhost:8081/test.aspx")); getPage.MakePageRequest(); string page = getPage.Results; Console.WriteLine(page); Assert.AreNotEqual(page.Length, 0); }
> > I am trying to get a webpage using a TcpSocket instead of a standard > > Webrequest. Initial, it works fine but after the 2 or 3 request the [quoted text clipped - 30 lines] > > Pete Peter Duniho - 29 Jun 2007 22:49 GMT > I am very sorry for the confusing. I was trying to be brief. Here is > the > class, and I renamed the socket to client for easyier reading. I am > using > TCPTrace at the moment to test this, and it gets and sends the proper > responses so I do not believe it is a Dos prevention of some kind. Well, I don't see anything obvious. As I mentioned, you really need to post a _complete_ sample of code that reliably reproduces the problem if anyone is to actually try to look at your specific problem. You've only posted the client side of things.
There do happen to be some odd things in your code, even without me inspecting each line carefully (which I didn't do). Nothing that seems like it would be causing the issue, but which are to me a sign of less-than-perfect code:
-- When you write to the NetworkStream, you use the length of the original string, not the byte array you're passing. In this case, they should be the same, but there are potential situations in which they wouldn't be. You should pass the length of the actual data you expect to send, not some other length that is related but not necessarily identical.
-- You check the CanWrite and CanRead properties of the stream, but don't do anything sensible if they aren't what you expect. One particular issue is that if CanWrite is false but CanRead is true, you skip trying to write but go ahead and try to read a response to something you never sent. That makes no sense. Presumably, both CanWrite and CanRead are true in this situation, so the oddity shouldn't affect you. But the code is wrong, nevertheless.
-- You are catching exceptions that you then do nothing with other than to throw them again. What's the point?
Have you tried not setting the timeout? It's not supposed to affect connecting, but maybe it does, or maybe having a timeout on a previous connection somehow interferes with the subsequent one. I admit, I'm just making wild guesses here...I don't see anything obvious. Even more reason for you to construct self-contained server/client sample code that reliably reproduces the problem. Just looking at the code doesn't suggest anything obvious, at least not to me.
Pete
Free MagazinesGet these publications absolutely FREE for up to 12 months. There are no hidden fees and no obligation. Simply choose a title, complete the application form and submit it. Read more ...
|
|
|