From owner-freebsd-hackers Sat Mar 8 23:15:27 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id XAA03685 for hackers-outgoing; Sat, 8 Mar 1997 23:15:27 -0800 (PST) Received: from crh.cl.msu.edu (crh.cl.msu.edu [35.8.1.24]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA03680 for ; Sat, 8 Mar 1997 23:15:25 -0800 (PST) Received: (from henrich@localhost) by crh.cl.msu.edu (8.8.5/8.8.4) id CAA02420; Sun, 9 Mar 1997 02:14:47 -0500 (EST) Date: Sun, 9 Mar 1997 02:14:47 -0500 (EST) From: Charles Henrich Message-Id: <199703090714.CAA02420@crh.cl.msu.edu> To: neal@pernet.net, freebsd-hackers@freebsd.org Subject: Re: freebsd as a news server? Newsgroups: lists.freebsd.hackers References: <5ftj7k$6v7$1@msunews.cl.msu.edu> X-Newsreader: NN version 6.5.0 #1 (NOV) Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In lists.freebsd.hackers you write: >> There's only 5 drives on it. I'm interested in figuring out >> what's wrong with the layout. I've got the striping factor set >> to 255 blocks (per a suggestion in the docs for ccd). I'll try >I've got the interleave set to 1k. I fiddled with the setting a little, >and found 1k appeared to give the best performance. If I remember Okay, this is incorrect. For a real high performance newserver you want the block size in the 65k range, as close to a cylinder group as possible. Certainly far more than 1k. You are trying to optimize transactions per second vs. raw bandwidth per second. 9gb disks are horribly slow last time I checked, if its what your stuck with, well, your stuck with it. >correctly, you want the "average" article to fit in one interleave. I You want the average newsgroup in one interleave :) We've found also that pushing the sync times back (sysctl parameter) from 30 seconds to 300 will give you nearly the performance increase as running async (if not exactly), and lend yourself to more stability in the absolute worst case. If running 2.2- you also want EXTRAVNODES up there in the 15000-30000 range depending on the size of your spool. -Crh -- Charles Henrich Michigan State University henrich@msu.edu http://pilot.msu.edu/~henrich