Return-Path: Delivered-To: apmail-httpd-dev-archive@httpd.apache.org Received: (qmail 22499 invoked by uid 500); 21 Nov 2001 18:37:53 -0000 Mailing-List: contact dev-help@httpd.apache.org; run by ezmlm Precedence: bulk Reply-To: dev@httpd.apache.org list-help: list-unsubscribe: list-post: Delivered-To: mailing list dev@httpd.apache.org Received: (qmail 22486 invoked from network); 21 Nov 2001 18:37:53 -0000 Content-Type: text/plain; charset="iso-8859-1" From: Ryan Bloom Reply-To: rbb@covalent.net Organization: Covalent Technologies To: dev@httpd.apache.org, Ian Holsman Subject: Re: core output filter buffering Date: Wed, 21 Nov 2001 10:37:05 -0800 X-Mailer: KMail [version 1.3] Cc: dev@httpd.apache.org References: <20011121181434.002D446DFD@koj.rkbloom.net> <1006367362.4469.28.camel@griffon.cnet.com> In-Reply-To: <1006367362.4469.28.camel@griffon.cnet.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-Id: <20011121183705.8F38C46DFD@koj.rkbloom.net> X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N On Wednesday 21 November 2001 10:29 am, Ian Holsman wrote: > On Wed, 2001-11-21 at 10:14, Ryan Bloom wrote: > > On Tuesday 20 November 2001 03:23 pm, Doug MacEachern wrote: > > > > Okay, I have looked at this again. I have a patch, but it is doing too > > much buffering now. :-( I'm hoping to fix this soon. However, there is > > a misunderstanding in the question below that I want to clear up. We try > > to buffer up to AP_MIN_BYTES_TO_WRITE in the core, never more than that. > > With that logic, the behavior you are seeing is correct. However, you > > are seeing the correct behavior because of the values that you chose. > > > > :-) > > Hey Ryan. > while your looking at buffering > is there any way to buffer the stuff up over brigades? > mod-include creates a seperate brigade per virtual include and you can > get lots of small packets being sent The core should be doing this for you. I think you are seeing the problem I described above. Namely, your SSI files are creating more than 16 brigades that combine to less than 8192 bytes. This is what causes the core to send the data too soon. I am about to commit a patch that stops that behavior. Can you test with the patch and let me know if it solves your problem too? If not, could you send me a test case? Ryan ______________________________________________________________ Ryan Bloom rbb@apache.org Covalent Technologies rbb@covalent.net --------------------------------------------------------------