From owner-freebsd-questions Sun Jun 25 14:17: 2 2000 Delivered-To: freebsd-questions@freebsd.org Received: from guru.mired.org (zoom0-125.telepath.com [216.14.0.125]) by hub.freebsd.org (Postfix) with SMTP id EBDB937B675 for ; Sun, 25 Jun 2000 14:16:51 -0700 (PDT) (envelope-from mwm@mired.org) Received: (qmail 53382 invoked by uid 100); 25 Jun 2000 21:16:49 -0000 From: Mike Meyer MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <14678.30401.351560.567647@guru.mired.org> Date: Sun, 25 Jun 2000 16:16:49 -0500 (CDT) To: questions@FreeBSD.ORG Subject: Updating mirrored web servers (Was: Equivalent in FreeBSD?) In-Reply-To: References: X-Mailer: VM 6.72 under 21.1 (patch 10) "Capitol Reef" XEmacs Lucid X-face: "5Mnwy%?j>IIV\)A=):rjWL~NB2aH[}Yq8Z=u~vJ`"(,&SiLvbbz2W`;h9L,Yg`+vb1>RG% *h+%X^n0EZd>TM8_IB;a8F?(Fb"lw'IgCoyM.[Lg#r\ Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Gabriel Ambuehl writes: > And then there's still the problem of the mirroring of the servers: > what solution is best to mirror them as close as possible to realtime? Perforce is a nice solution (though it's not free). You need at least two branches to the web server code, though. Whatever you want for development & testing, and then a "production" branch. Your production web servers are configured so they do a "p4 sync" and it copies the latest production code to the proper places on the web server. This is cheap because Perforce keeps (on the perforce server) a database of what is on each client. In fact, it's so cheap that having a script that does it once a minute won't noticably load even a 486. With lots of web servers, doing it frequently is probably a *good* idea. Doing it often keeps the load on the perforce server closer to level, as it updates small bits frequenty rather than large bits at longer intervals. The end result is that when someone commits code to the production branch, all the web servers are udpated in a matter of minutes. What's really slick is that if your site search engine (you do have a site search engine, don't you?) has an "udpate this page" hook, you can feed the list of updated files output by "p4 sync" - after proper mastication - to the site search engine, and have real-time udpates to that as well.