Return-Path: Delivered-To: apmail-apr-dev-archive@apr.apache.org Received: (qmail 4484 invoked by uid 500); 24 Sep 2001 15:31:21 -0000 Mailing-List: contact dev-help@apr.apache.org; run by ezmlm Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: Delivered-To: mailing list dev@apr.apache.org Received: (qmail 4443 invoked from network); 24 Sep 2001 15:31:21 -0000 X-Authentication-Warning: rdu88-250-106.nc.rr.com: trawick set sender to trawick@attglobal.net using -f Sender: trawick@rdu88-250-106.nc.rr.com To: Justin Erenkrantz Cc: dev@apr.apache.org Subject: Re: [PATCH] Switch to pthread as the default interprocess lock... References: <20010922122558.B6756@ebuilt.com> From: Jeff Trawick Date: 24 Sep 2001 11:34:39 -0400 In-Reply-To: <20010922122558.B6756@ebuilt.com> Message-ID: Lines: 35 User-Agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/20.3 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N Justin Erenkrantz writes: > I've brought this up before and I'll bring it up again. =-) > > Ian has seen about a ~60rps improvement using AcceptMutex pthread > versus fcntl() on Solaris with httpd. Also, fcntl has scalability > issues from what I can tell on Solaris (as described in STATUS) > that cause fatal errors in httpd. > > I know that the AIX guys use pthread as well. > > Is there any platform that meets the criteria for our pthread > interprocess implementation where it is *not* better than fcntl()? That is the wrong question for deciding whether or not to set the default to pthread IMHO. There is empirical evidence (lots of happy 1.3 users) that fcntl is a good overall choice. If/when we find that a certain platform should use a different mechanism, I think we should specify *for that platform only* to use a different mechanism rather than consider changing the default on multiple platforms. pthread isn't even the best choice across the board on Solaris*. How are we going to be able to evaluate whether or not it should be used on platforms we don't use/test extensively? *I don't mean to infer that fcntl is the best choice across the board on Solaris. Instead, I mean that certain Solaris configurations shouldn't be using pthread. -- Jeff Trawick | trawick@attglobal.net | PGP public key at web site: http://www.geocities.com/SiliconValley/Park/9289/ Born in Roswell... married an alien...