Subject: Re: Strange Nagle effect on 1.6.2
To: Alan Barrett <email@example.com>
From: David Laight <firstname.lastname@example.org>
Date: 05/30/2005 21:46:44
On Mon, May 30, 2005 at 08:52:16PM +0200, Alan Barrett wrote:
> If the immediate sending implied by the TCP PSH flag took precedence
> over the delated sending implied by the Nagle algorithm, then this
> problem would not arise. Is there any reason not to let the PSH flag
> take precedence?
IIRC the PSH flag is/was intended to ensure that the sent data is
delivered to the application without further delay. A system doing
forwarding (at the TCP level) could otherwise wait for further data
and coalesce the data into a single TCP message.
In particular the PSH flag isn't exposed to the application itself by
any APIs I know of. So the application cannot use it to pervert the
cause of Nagle on a per-send basis.
David Laight: email@example.com