Subject: RE: Packet drop on full socket problem

RE: Packet drop on full socket problem

From: Daniel Stenberg <daniel_at_haxx.se>
Date: Thu, 7 Oct 2010 11:42:37 +0200 (CEST)

On Thu, 7 Oct 2010, Thomas Rauscher wrote:

>>> * _libssh2_transport_write then returns "wrote" (12k) to
>> the application.
>>
>> Right, as it did in fact successfully send away 12K.
>
> I think that this is the core problem. The code has sent away 12k, but
> failed to send the rest. It saves the 12k-Buffer to odata and the upper
> level function _libssh2_channel_write throws it away again. This means that
> this buffer is missing in the SSH stream.

When libssh2_channel_write_ex() returns 12K and thus informs the application
that 12K data was indeed sent, the appliction will advance its pointer and the
next call will be 12K further into the data.

Thus, the application will deliver the rest of the data (again) and libssh2
won't need to have it buffered anywhere. That buffer is only needed and used
as long as it truly and only gets EAGAIN back from the socket.

A short send() is very common for anything except very small files when sent
over a slower network, so I don't think it is "just" that's required to
trigger this.

-- 
  / daniel.haxx.se
_______________________________________________
libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel
Received on 2010-10-07