[issue12498] asyncore.dispatcher_with_send, disconnection problem + miss-conception

[Charles-François Natali]

Charles-François Natali <neologix at free.fr> added the comment:

Hello,

> Actually the class asyncore.dispatcher_with_send do not handle properly
> disconnection. When the endpoint shutdown his sending part of the socket,
> but keep the socket open in reading, the current implementation of
> dispatcher_with_send will close the socket without sending pending data.

Yes, that's a common problem with "naive" networking code.

> Note also that this class try to initiate a send even if the socket is maybe
> not ready for writing:
>
> Here's a simple fix:
>     def send(self, data):
>         if self.debug:
>             self.log_info('sending %s' % repr(data))
>         self.out_buffer = self.out_buffer + data
> -       self.initiate_send()
>

Hum, I'm not sure about this.
dispatcher is just a thin wrapper around the underlying socket, so the
semantic of `send` should be the same as `socket.send`, i.e. just call
the send(2) syscall. I think it's the application's reponsibility to
check that the socket is indeed writable, and to cope with potential
failures (e.g. EAGAIN or ENOTCONN).

> Last but not last, the buffer size of each socket send are way to small
> (512, a third of an usual TCP frame). Here's the code with a bumped value:

Indeed, 1/3 of the ethernet MTU is quite small.

Would you like to submit a patch?

----------
nosy: +neologix

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue12498>
_______________________________________