From owner-freebsd-isp Sun Nov 10 11:39:38 1996 Return-Path: owner-isp Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id LAA28510 for isp-outgoing; Sun, 10 Nov 1996 11:39:38 -0800 (PST) Received: from mx.serv.net (mx.serv.net [199.201.191.10]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id LAA28502 for ; Sun, 10 Nov 1996 11:39:33 -0800 (PST) Received: from MindBender.serv.net by mx.serv.net (8.7.5/SERV Revision: 2.30) id LAA23828; Sun, 10 Nov 1996 11:39:30 -0800 (PST) Received: from localhost.HeadCandy.com (michaelv@localhost.HeadCandy.com [127.0.0.1]) by MindBender.serv.net (8.7.5/8.7.3) with SMTP id LAA23455; Sun, 10 Nov 1996 11:39:21 -0800 (PST) Message-Id: <199611101939.LAA23455@MindBender.serv.net> X-Authentication-Warning: MindBender.serv.net: Host michaelv@localhost.HeadCandy.com [127.0.0.1] didn't use HELO protocol To: hgoldste@bbs.mpcs.com cc: freebsd-isp@freebsd.org Subject: Re: Best mount options, tunefs for newsserver In-reply-to: Your message of Sun, 10 Nov 96 07:22:42 -0500. <199611101222.HAA19687@bbs.mpcs.com> Date: Sun, 10 Nov 1996 11:39:21 -0800 From: "Michael L. VanLoon -- HeadCandy.com" Sender: owner-isp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > : >Trying to tweak the newsserver here. I have the slices that are > : >hammered all mounted async (are they really working async? is there a > : >way to tell?). Any other useful mount options for a local fs? > : I think so. I run my NetBSD box this way, and it's pretty obvious > : sometimes that it's async. For example, copy a big directory of lots > : of files (like /usr/src/) to somewhere, sync it and let it settle, > : then rm -rf it. It should finish really quickly, then a burst of disk > : activity when sync time comes. >Hmmm this is very interesting. I did see a delay until sync time came >but what's really interesting is the huge amount of disk access that >took place asynchronously. > >The time difference between rm -r'ing /usr/src/sys (copied somewhere >else :) ) on a sync vs async mounted filesystem was not discernable. >Each rm -r run took about 10 seconds; the async run included a little >burst of flushing activity about 2 seconds after the shell prompt >returned. Interesting. It's a pretty obvious difference when I do it on NetBSD (and I don't think there should be a big difference in behavior between the BSDs in this particular case). >I think there should be an adequate amount of cache available on a >32mb system with only myself as the user. I suppose it's possible you could over-run the cache if the stuff you were rm'ing was too big. >Any suggestions on a definitive way to find out the extent of caching >going on? Is the fact that the slice I tested on is marked NFS >exported (but not being actively exported) a factor here? >Should I be asking this on the fs mailing list? I'll have to defer this to someone more knowledgable about the internals of FreeBSD... ----------------------------------------------------------------------------- Michael L. VanLoon michaelv@MindBender.serv.net --< Free your mind and your machine -- NetBSD free un*x >-- NetBSD working ports: 386+PC, Mac 68k, Amiga, Atari 68k, HP300, Sun3, Sun4/4c/4m, DEC MIPS, DEC Alpha, PC532, VAX, MVME68k, arm32... NetBSD ports in progress: PICA, others... -----------------------------------------------------------------------------