提高asio超时

发布于 2024-11-07 18:34:05 字数 957 浏览 1 评论 0原文

可能的重复：
如何在 boost 中设置阻塞套接字的超时亚洲？

我之前读过一些关于超时的条目，但我不明白。

我想要一个定义的连接超时。连接代码如下所示：

try{
  boost::asio::ip::tcp::resolver              resolver(m_ioService);
  boost::asio::ip::tcp::resolver::query       query(link.get_host(), link.get_scheme());
  boost::asio::ip::tcp::resolver::iterator    endpoint_iterator = resolver.resolve(query);
  boost::asio::ip::tcp::resolver::iterator    end;
  boost::system::error_code                   error   =   boost::asio::error::host_not_found;

  while (error && endpoint_iterator != end)
   {
    m_socket.close();
    m_socket.connect(*endpoint_iterator++, error);
   }
}

我也想要一个读取超时。

我使用 boost::asio::read_until(m_socket, response, "\r\n"); 来读取标头。

是否可以设置简单的超时？

原文

Possible Duplicate:
How to set a timeout on blocking sockets in boost asio?

I read some of the entries before about the timeout but I don't understand.

I want a defined timeout for the connection.
the connect code looks like:

try{
  boost::asio::ip::tcp::resolver              resolver(m_ioService);
  boost::asio::ip::tcp::resolver::query       query(link.get_host(), link.get_scheme());
  boost::asio::ip::tcp::resolver::iterator    endpoint_iterator = resolver.resolve(query);
  boost::asio::ip::tcp::resolver::iterator    end;
  boost::system::error_code                   error   =   boost::asio::error::host_not_found;

  while (error && endpoint_iterator != end)
   {
    m_socket.close();
    m_socket.connect(*endpoint_iterator++, error);
   }
}

also I want a read timeout.

I use boost::asio::read_until(m_socket, response, "\r\n"); for read the header.

is it possible to set SIMPLE a timeout?

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

自此以后，行同陌路 2024-11-14 18:34:05

首先，我相信您应该始终使用异步方法，因为它们更好，并且您的设计只会从反应器模式方法中受益。
在糟糕的情况下，您很着急并且正在进行原型设计，同步方法可能会很有用。在这种情况下，我确实同意你的观点，如果没有任何超时支持，它们就无法在现实世界中使用。

我所做的非常简单：

void HttpClientImpl::configureSocketTimeouts(boost::asio::ip::tcp::socket& socket)
{
#if defined OS_WINDOWS
    int32_t timeout = 15000;
    setsockopt(socket.native(), SOL_SOCKET, SO_RCVTIMEO, (const char*)&timeout, sizeof(timeout));
    setsockopt(socket.native(), SOL_SOCKET, SO_SNDTIMEO, (const char*)&timeout, sizeof(timeout));
#else
    struct timeval tv;
    tv.tv_sec  = 15; 
    tv.tv_usec = 0;         
    setsockopt(socket.native(), SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof(tv));
    setsockopt(socket.native(), SOL_SOCKET, SO_SNDTIMEO, &tv, sizeof(tv));
#endif
}

根据 OS_WINDOWS 宏，上面的代码可以在 Windows、Linux 和 MAC OS 上运行。

Fist of all I believe that you should ALWAYS use the async methods since they are better and your design will only benefit from a reactor pattern approach.
In the bad case that you're in a hurry and you're kind of prototyping, the sync methods can be useful. In this case I do agree with you that without any timeout support, they cannot be used in the real world.

What I did was very simple:

void HttpClientImpl::configureSocketTimeouts(boost::asio::ip::tcp::socket& socket)
{
#if defined OS_WINDOWS
    int32_t timeout = 15000;
    setsockopt(socket.native(), SOL_SOCKET, SO_RCVTIMEO, (const char*)&timeout, sizeof(timeout));
    setsockopt(socket.native(), SOL_SOCKET, SO_SNDTIMEO, (const char*)&timeout, sizeof(timeout));
#else
    struct timeval tv;
    tv.tv_sec  = 15; 
    tv.tv_usec = 0;         
    setsockopt(socket.native(), SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof(tv));
    setsockopt(socket.native(), SOL_SOCKET, SO_SNDTIMEO, &tv, sizeof(tv));
#endif
}

The code above works both on windows and on Linux and on MAC OS, according to the OS_WINDOWS macro.

回复收藏 0 原文