HttpWebResponse get stuck while running in a loop - c#

I build this method (c#) in order to receive the HTTP response status code from an URL.
whene I run this method ones it's works fine, but when I run it in a loop, the third time its stuck. any clue??
public static string isAlive(string url)
{
Console.WriteLine("start: Is Alive Test");
WebRequest request = WebRequest.Create(url);
try
{
HttpWebResponse response = (HttpWebResponse)request.GetResponse();
return Convert.ToString((int)response.StatusCode);
}
catch(WebException ex)
{
HttpWebResponse res = (HttpWebResponse)ex.Response;
return Convert.ToString((int)res.StatusCode);
}
}
the loop
for (int i = 0; i < 5; i++)
{
string a = isAlive("https://www.yahoo.com/");
Console.WriteLine(a);
}

You're not calling Dispose on the HttpWebResponse object, which means that the connection is still lying around. If you change your code to the following:
public static string isAlive(string url)
{
Console.WriteLine("start: Is Alive Test");
WebRequest request = WebRequest.Create(url);
try
{
using(HttpWebResponse response = (HttpWebResponse)request.GetResponse())
{
return Convert.ToString((int)response.StatusCode);
}
}
catch(WebException ex)
{
using(HttpWebResponse res = (HttpWebResponse)ex.Response)
{
return Convert.ToString((int)res.StatusCode);
}
}
}
the using statement will implicitly call Dispose for you, which will close the connection.
The reason your code is halting after the second iteration is because .Net has a built in maximum number of connections it will open to a website, which is by default 2. This is controlled by System.Net.ServicePointManager.DefaultConnectionLimit which you can increase should you need to.

You need to wrap HttpWebResponse var into using statement because it's disposable
Before checking ex.Response.StatusCode make sure that ex.Status is a ProtocolError
And also consider making your method asynchronous for performance considerations
Since your method is returning a status code, there might be a better name for it than isAlive
Sample:
public static async Task<string> GetStatusCode(string url)
{
var request = (HttpWebRequest)WebRequest.Create(url);
try
{
using (var response = (HttpWebResponse)await request.GetResponseAsync())
{
return response.StatusCode.ToString();
}
}
catch (WebException ex)
{
return ex.Status == WebExceptionStatus.ProtocolError ?
((HttpWebResponse)e.Response).StatusCode.ToString() : null;
}
}

It might have to do with you not closing the HttpWebResponse.
Add a finally to that try catch which closes the response.
Also close the WebException response within the catch.

Use "using" and it will work well.
using (HttpWebResponse response = (HttpWebResponse)request.GetResponse())
{
return Convert.ToString((int)response.StatusCode);
}

Related

C# GetResponse() How to timeout dynamically created url?

This is my method, to which I am passing the url to check if it's active.
The link is being activated on the wowza service so it takes some time until it's "alive"
GetResponse is returning the 404 Error because the url is not reached.
Is there a way to get the timeout instead of 404 error if the url is not alive after specified time?
public async Task<IActionResult> GetLinkIsAlive([FromQuery] string url, [FromQuery] int timeout)
{
HttpWebResponse webResponse;
try
{
HttpWebRequest webRequest = (HttpWebRequest)WebRequest.Create(url);
webRequest.Timeout = timeout;
webRequest.Method = "GET";
webResponse = webRequest.GetResponse() as HttpWebResponse;
return Ok(webResponse.StatusCode);
}
catch (WebException ex)
{
return BadRequest(ex.Message);
}
}
You can use connection pooling.
It’s using IHttpClientFactory that helps to maintain the pooling and lifetime of clients
In your startup class :
services.AddHttpClient<NameofyourService, NameofyourService>()
.AddTransientHttpErrorPolicy(
p => p.WaitAndRetryAsync(new[]
{
TimeSpan.FromSeconds(1),
TimeSpan.FromSeconds(5),
TimeSpan.FromSeconds(10)
}));
It requires Microsoft.Extensions.Http.Polly
You need to use HttpClient for your service. All added configurations will apply automatically.
My solution to this was calling the link in the while loop every 1s and await the whole Task.
private async Task<bool> IsLiveStreamAlive(string streamUrl, int retriesCount = 30)
{
try
{
bool res = false;
HttpClient client = new HttpClient();
Uri uri = new Uri(streamUrl);
HttpWebRequest webRequest = (HttpWebRequest)WebRequest.Create(uri);
var request = new HttpRequestMessage(HttpMethod.Get, uri);
while (retriesCount > 1)
{
await Task.Delay(1000);
retriesCount--;
HttpResponseMessage httpResponse = await client.GetAsync(uri);
res = httpResponse.StatusCode == HttpStatusCode.OK ? true : false;
if (res)
{
Log.Info(string.Format("Stream alive: {0}", streamUrl));
break;
}
}
return res;
}
catch (WebException ex)
{
Log.Error(ex);
}
return false;
}

How to read status code from response body of c# webclient request [duplicate]

I am using the WebClient class to post some data to a web form. I would like to get the response status code of the form submission. So far I've found out how to get the status code if there is a exception
Catch wex As WebException
If TypeOf wex.Response Is HttpWebResponse Then
msgbox(DirectCast(wex.Response, HttpWebResponse).StatusCode)
End If
However if the form is submitted successfully and no exception is thrown then I won't know the status code(200,301,302,...)
Is there some way to get the status code when there is no exceptions thrown?
PS: I prefer not to use httpwebrequest/httpwebresponse
You can check if the error is of type WebException and then inspect the response code;
if (e.Error.GetType().Name == "WebException")
{
WebException we = (WebException)e.Error;
HttpWebResponse response = (System.Net.HttpWebResponse)we.Response;
if (response.StatusCode==HttpStatusCode.NotFound)
System.Diagnostics.Debug.WriteLine("Not found!");
}
or
try
{
// send request
}
catch (WebException e)
{
// check e.Status as above etc..
}
There is a way to do it using reflection. It works with .NET 4.0. It accesses a private field and may not work in other versions of .NET without modifications.
I have no idea why Microsoft did not expose this field with a property.
private static int GetStatusCode(WebClient client, out string statusDescription)
{
FieldInfo responseField = client.GetType().GetField("m_WebResponse", BindingFlags.Instance | BindingFlags.NonPublic);
if (responseField != null)
{
HttpWebResponse response = responseField.GetValue(client) as HttpWebResponse;
if (response != null)
{
statusDescription = response.StatusDescription;
return (int)response.StatusCode;
}
}
statusDescription = null;
return 0;
}
If you are using .Net 4.0 (or less):
class BetterWebClient : WebClient
{
private WebRequest _Request = null;
protected override WebRequest GetWebRequest(Uri address)
{
this._Request = base.GetWebRequest(address);
if (this._Request is HttpWebRequest)
{
((HttpWebRequest)this._Request).AllowAutoRedirect = false;
}
return this._Request;
}
public HttpStatusCode StatusCode()
{
HttpStatusCode result;
if (this._Request == null)
{
throw (new InvalidOperationException("Unable to retrieve the status
code, maybe you haven't made a request yet."));
}
HttpWebResponse response = base.GetWebResponse(this._Request)
as HttpWebResponse;
if (response != null)
{
result = response.StatusCode;
}
else
{
throw (new InvalidOperationException("Unable to retrieve the status
code, maybe you haven't made a request yet."));
}
return result;
}
}
If you are using .Net 4.5.X or newer, switch to HttpClient:
var response = await client.GetAsync("http://www.contoso.com/");
var statusCode = response.StatusCode;
Tried it out. ResponseHeaders do not include status code.
If I'm not mistaken, WebClient is capable of abstracting away multiple distinct requests in a single method call (e.g. correctly handling 100 Continue responses, redirects, and the like). I suspect that without using HttpWebRequest and HttpWebResponse, a distinct status code may not be available.
It occurs to me that, if you are not interested in intermediate status codes, you can safely assume the final status code is in the 2xx (successful) range, otherwise, the call would not be successful.
The status code unfortunately isn't present in the ResponseHeaders dictionary.
Erik's answer doesn't work on Windows Phone as is. The following does:
class WebClientEx : WebClient
{
private WebResponse m_Resp = null;
protected override WebResponse GetWebResponse(WebRequest Req, IAsyncResult ar)
{
try
{
this.m_Resp = base.GetWebResponse(request);
}
catch (WebException ex)
{
if (this.m_Resp == null)
this.m_Resp = ex.Response;
}
return this.m_Resp;
}
public HttpStatusCode StatusCode
{
get
{
if (m_Resp != null && m_Resp is HttpWebResponse)
return (m_Resp as HttpWebResponse).StatusCode;
else
return HttpStatusCode.OK;
}
}
}
At least it does when using OpenReadAsync; for other xxxAsync methods, careful testing would be highly recommended. The framework calls GetWebResponse somewhere along the code path; all one needs to do is capture and cache the response object.
The fallback code is 200 in this snippet because genuine HTTP errors - 500, 404, etc - are reported as exceptions anyway. The purpose of this trick is to capture non-error codes, in my specific case 304 (Not modified). So the fallback assumes that if the status code is somehow unavailable, at least it's a non-erroneous one.
You should use
if (e.Status == WebExceptionStatus.ProtocolError)
{
HttpWebResponse response = (HttpWebResponse)ex.Response;
if (response.StatusCode == HttpStatusCode.NotFound)
System.Diagnostics.Debug.WriteLine("Not found!");
}
This is what I use for expanding WebClient functionality. StatusCode and StatusDescription will always contain the most recent response code/description.
/// <summary>
/// An expanded web client that allows certificate auth and
/// the retrieval of status' for successful requests
/// </summary>
public class WebClientCert : WebClient
{
private X509Certificate2 _cert;
public WebClientCert(X509Certificate2 cert) : base() { _cert = cert; }
protected override WebRequest GetWebRequest(Uri address)
{
HttpWebRequest request = (HttpWebRequest)base.GetWebRequest(address);
if (_cert != null) { request.ClientCertificates.Add(_cert); }
return request;
}
protected override WebResponse GetWebResponse(WebRequest request)
{
WebResponse response = null;
response = base.GetWebResponse(request);
HttpWebResponse baseResponse = response as HttpWebResponse;
StatusCode = baseResponse.StatusCode;
StatusDescription = baseResponse.StatusDescription;
return response;
}
/// <summary>
/// The most recent response statusCode
/// </summary>
public HttpStatusCode StatusCode { get; set; }
/// <summary>
/// The most recent response statusDescription
/// </summary>
public string StatusDescription { get; set; }
}
Thus you can do a post and get result via:
byte[] response = null;
using (WebClientCert client = new WebClientCert())
{
response = client.UploadValues(postUri, PostFields);
HttpStatusCode code = client.StatusCode;
string description = client.StatusDescription;
//Use this information
}
Just in case someone else needs an F# version of the above described hack.
open System
open System.IO
open System.Net
type WebClientEx() =
inherit WebClient ()
[<DefaultValue>] val mutable m_Resp : WebResponse
override x.GetWebResponse (req: WebRequest ) =
x.m_Resp <- base.GetWebResponse(req)
(req :?> HttpWebRequest).AllowAutoRedirect <- false;
x.m_Resp
override x.GetWebResponse (req: WebRequest , ar: IAsyncResult ) =
x.m_Resp <- base.GetWebResponse(req, ar)
(req :?> HttpWebRequest).AllowAutoRedirect <- false;
x.m_Resp
member x.StatusCode with get() : HttpStatusCode =
if not (obj.ReferenceEquals (x.m_Resp, null)) && x.m_Resp.GetType() = typeof<HttpWebResponse> then
(x.m_Resp :?> HttpWebResponse).StatusCode
else
HttpStatusCode.OK
let wc = new WebClientEx()
let st = wc.OpenRead("http://www.stackoverflow.com")
let sr = new StreamReader(st)
let res = sr.ReadToEnd()
wc.StatusCode
sr.Close()
st.Close()
You should be able to use the "client.ResponseHeaders[..]" call, see this link for examples of getting stuff back from the response
You can try this code to get HTTP status code from WebException or from OpenReadCompletedEventArgs.Error. It works in Silverlight too because SL does not have WebExceptionStatus.ProtocolError defined.
HttpStatusCode GetHttpStatusCode(System.Exception err)
{
if (err is WebException)
{
WebException we = (WebException)err;
if (we.Response is HttpWebResponse)
{
HttpWebResponse response = (HttpWebResponse)we.Response;
return response.StatusCode;
}
}
return 0;
}

WebRequest and protocol agnostic URL

In a web application I have to first check if an image exists and then display this image or a dummy image.
I use the following code and it works for URLS like:
"http://www.somedomain.com/niceimage.png"
"https://www.somedomain.com/niceimage.png".
public virtual bool WebResourceExists(string url)
{
WebHeaderCollection headers = null;
WebResponse response = null;
try
{
WebRequest request = WebRequest.Create(url);
request.Method = "HEAD";
response = request.GetResponse();
headers = response.Headers;
bool result = int.Parse(headers["Content-Length"]) > 0;
return result;
}
catch (System.Net.WebException)
{
return false;
}
catch (Exception e)
{
_log.Error(e);
return false;
}
finally
{
if (response != null)
{
response.Close();
}
}
}
In some places the the method is called with protocol agnostic urls like "//www.somedomain.com/niceimage.png".
There is an exception thrown for such urls:
System.InvalidCastException: Unable to cast object of type
'System.Net.FileWebRequest' to type 'System.Net.HttpWebRequest'
Is there a way to use protocol agnostic urls other then just prepending "http:" to the url?
Protocol-agnostic URLs are resolved by the browser using the current protocol, and are used to avoid making HTTP requests from an HTTPS page.
Code executing on the server doesn't really have a concept of a "current protocol". Whilst ASP.NET can determine whether the current request was issued over HTTP or HTTPS, the WebRequest classes are not restricted to ASP.NET applications, so they cannot rely on this.
You will need to specify the protocol. Whether you use HTTP or HTTPS will depend on whether you're concerned about third-parties eavesdropping on the connection between your server and "www.somedomain.com".
What about a two step process, check for the http version, if it doesn't exist check for https. I've quickly hacked together a basic example of how this could work but am not able to properly test and check it so it might need some tidying up/refactoring!
public virtual bool WebResourceExists(string url)
{
WebHeaderCollection headers = null;
WebResponse response = null;
try
{
if (url.StartsWith(#"//") {
url = "http:";
}
WebRequest request = WebRequest.Create(url);
request.Method = "HEAD";
response = request.GetResponse();
headers = response.Headers;
bool result = int.Parse(headers["Content-Length"]) > 0;
return result;
}
catch (System.Net.WebException)
{
if (url.StartsWith(#"http://") {
url = url.Replace("http://","https://");
} else {
return false;
}
try {
WebRequest request = WebRequest.Create(url);
request.Method = "HEAD";
response = request.GetResponse();
headers = response.Headers;
bool result = int.Parse(headers["Content-Length"]) > 0;
return result;
}
catch (System.Net.WebException)
{
return false;
}
}
catch (Exception e)
{
_log.Error(e);
return false;
}
finally
{
if (response != null)
{
response.Close();
}
}
}

handling httpwebrequest.getresponse exception inside timer function

i'm using httpwebrequest.getresponse() inside function with timer. Timer calls this function every few seconds. However sometimes web server refuses to response due to quick request and response because of timer.
I need to make sure that function continues execution with timer and exception is handled properly.
or
what is the best way to handle http web request and web response for a function inside timer?
Handle the exception the usual way, then?
try {
// Send your request
} catch(WebException ex) {
// It failed
}
while(response==null)
{
httpWebRequest = (HttpWebRequest)WebRequest.Create(sURL + param);
httpWebRequest.Method = WebRequestMethods.Http.Post;
httpWebRequest.Accept = "application/xml";
httpWebRequest.ContentLength = 0;
try
{
response = (HttpWebResponse)httpWebRequest.GetResponse();
if (response.StatusCode == HttpStatusCode.OK && response != null)
{
streamReader = new StreamReader(response.GetResponseStream());
doc = XDocument.Load(streamReader);
break;
}
}
catch (WebException exx)
{
Console.WriteLine("Trying to reconnect with Web Server");
Thread.Sleep(2000);
}
}

How to get status code from webclient?

I am using the WebClient class to post some data to a web form. I would like to get the response status code of the form submission. So far I've found out how to get the status code if there is a exception
Catch wex As WebException
If TypeOf wex.Response Is HttpWebResponse Then
msgbox(DirectCast(wex.Response, HttpWebResponse).StatusCode)
End If
However if the form is submitted successfully and no exception is thrown then I won't know the status code(200,301,302,...)
Is there some way to get the status code when there is no exceptions thrown?
PS: I prefer not to use httpwebrequest/httpwebresponse
You can check if the error is of type WebException and then inspect the response code;
if (e.Error.GetType().Name == "WebException")
{
WebException we = (WebException)e.Error;
HttpWebResponse response = (System.Net.HttpWebResponse)we.Response;
if (response.StatusCode==HttpStatusCode.NotFound)
System.Diagnostics.Debug.WriteLine("Not found!");
}
or
try
{
// send request
}
catch (WebException e)
{
// check e.Status as above etc..
}
There is a way to do it using reflection. It works with .NET 4.0. It accesses a private field and may not work in other versions of .NET without modifications.
I have no idea why Microsoft did not expose this field with a property.
private static int GetStatusCode(WebClient client, out string statusDescription)
{
FieldInfo responseField = client.GetType().GetField("m_WebResponse", BindingFlags.Instance | BindingFlags.NonPublic);
if (responseField != null)
{
HttpWebResponse response = responseField.GetValue(client) as HttpWebResponse;
if (response != null)
{
statusDescription = response.StatusDescription;
return (int)response.StatusCode;
}
}
statusDescription = null;
return 0;
}
If you are using .Net 4.0 (or less):
class BetterWebClient : WebClient
{
private WebRequest _Request = null;
protected override WebRequest GetWebRequest(Uri address)
{
this._Request = base.GetWebRequest(address);
if (this._Request is HttpWebRequest)
{
((HttpWebRequest)this._Request).AllowAutoRedirect = false;
}
return this._Request;
}
public HttpStatusCode StatusCode()
{
HttpStatusCode result;
if (this._Request == null)
{
throw (new InvalidOperationException("Unable to retrieve the status
code, maybe you haven't made a request yet."));
}
HttpWebResponse response = base.GetWebResponse(this._Request)
as HttpWebResponse;
if (response != null)
{
result = response.StatusCode;
}
else
{
throw (new InvalidOperationException("Unable to retrieve the status
code, maybe you haven't made a request yet."));
}
return result;
}
}
If you are using .Net 4.5.X or newer, switch to HttpClient:
var response = await client.GetAsync("http://www.contoso.com/");
var statusCode = response.StatusCode;
Tried it out. ResponseHeaders do not include status code.
If I'm not mistaken, WebClient is capable of abstracting away multiple distinct requests in a single method call (e.g. correctly handling 100 Continue responses, redirects, and the like). I suspect that without using HttpWebRequest and HttpWebResponse, a distinct status code may not be available.
It occurs to me that, if you are not interested in intermediate status codes, you can safely assume the final status code is in the 2xx (successful) range, otherwise, the call would not be successful.
The status code unfortunately isn't present in the ResponseHeaders dictionary.
Erik's answer doesn't work on Windows Phone as is. The following does:
class WebClientEx : WebClient
{
private WebResponse m_Resp = null;
protected override WebResponse GetWebResponse(WebRequest Req, IAsyncResult ar)
{
try
{
this.m_Resp = base.GetWebResponse(request);
}
catch (WebException ex)
{
if (this.m_Resp == null)
this.m_Resp = ex.Response;
}
return this.m_Resp;
}
public HttpStatusCode StatusCode
{
get
{
if (m_Resp != null && m_Resp is HttpWebResponse)
return (m_Resp as HttpWebResponse).StatusCode;
else
return HttpStatusCode.OK;
}
}
}
At least it does when using OpenReadAsync; for other xxxAsync methods, careful testing would be highly recommended. The framework calls GetWebResponse somewhere along the code path; all one needs to do is capture and cache the response object.
The fallback code is 200 in this snippet because genuine HTTP errors - 500, 404, etc - are reported as exceptions anyway. The purpose of this trick is to capture non-error codes, in my specific case 304 (Not modified). So the fallback assumes that if the status code is somehow unavailable, at least it's a non-erroneous one.
You should use
if (e.Status == WebExceptionStatus.ProtocolError)
{
HttpWebResponse response = (HttpWebResponse)ex.Response;
if (response.StatusCode == HttpStatusCode.NotFound)
System.Diagnostics.Debug.WriteLine("Not found!");
}
This is what I use for expanding WebClient functionality. StatusCode and StatusDescription will always contain the most recent response code/description.
/// <summary>
/// An expanded web client that allows certificate auth and
/// the retrieval of status' for successful requests
/// </summary>
public class WebClientCert : WebClient
{
private X509Certificate2 _cert;
public WebClientCert(X509Certificate2 cert) : base() { _cert = cert; }
protected override WebRequest GetWebRequest(Uri address)
{
HttpWebRequest request = (HttpWebRequest)base.GetWebRequest(address);
if (_cert != null) { request.ClientCertificates.Add(_cert); }
return request;
}
protected override WebResponse GetWebResponse(WebRequest request)
{
WebResponse response = null;
response = base.GetWebResponse(request);
HttpWebResponse baseResponse = response as HttpWebResponse;
StatusCode = baseResponse.StatusCode;
StatusDescription = baseResponse.StatusDescription;
return response;
}
/// <summary>
/// The most recent response statusCode
/// </summary>
public HttpStatusCode StatusCode { get; set; }
/// <summary>
/// The most recent response statusDescription
/// </summary>
public string StatusDescription { get; set; }
}
Thus you can do a post and get result via:
byte[] response = null;
using (WebClientCert client = new WebClientCert())
{
response = client.UploadValues(postUri, PostFields);
HttpStatusCode code = client.StatusCode;
string description = client.StatusDescription;
//Use this information
}
Just in case someone else needs an F# version of the above described hack.
open System
open System.IO
open System.Net
type WebClientEx() =
inherit WebClient ()
[<DefaultValue>] val mutable m_Resp : WebResponse
override x.GetWebResponse (req: WebRequest ) =
x.m_Resp <- base.GetWebResponse(req)
(req :?> HttpWebRequest).AllowAutoRedirect <- false;
x.m_Resp
override x.GetWebResponse (req: WebRequest , ar: IAsyncResult ) =
x.m_Resp <- base.GetWebResponse(req, ar)
(req :?> HttpWebRequest).AllowAutoRedirect <- false;
x.m_Resp
member x.StatusCode with get() : HttpStatusCode =
if not (obj.ReferenceEquals (x.m_Resp, null)) && x.m_Resp.GetType() = typeof<HttpWebResponse> then
(x.m_Resp :?> HttpWebResponse).StatusCode
else
HttpStatusCode.OK
let wc = new WebClientEx()
let st = wc.OpenRead("http://www.stackoverflow.com")
let sr = new StreamReader(st)
let res = sr.ReadToEnd()
wc.StatusCode
sr.Close()
st.Close()
You should be able to use the "client.ResponseHeaders[..]" call, see this link for examples of getting stuff back from the response
You can try this code to get HTTP status code from WebException or from OpenReadCompletedEventArgs.Error. It works in Silverlight too because SL does not have WebExceptionStatus.ProtocolError defined.
HttpStatusCode GetHttpStatusCode(System.Exception err)
{
if (err is WebException)
{
WebException we = (WebException)err;
if (we.Response is HttpWebResponse)
{
HttpWebResponse response = (HttpWebResponse)we.Response;
return response.StatusCode;
}
}
return 0;
}

Categories