From owner-freebsd-isp Mon Jun 7 10: 5:18 1999 Delivered-To: freebsd-isp@freebsd.org Received: from home.dragondata.com (home.dragondata.com [204.137.237.2]) by hub.freebsd.org (Postfix) with ESMTP id B909A151C7 for ; Mon, 7 Jun 1999 10:05:06 -0700 (PDT) (envelope-from toasty@home.dragondata.com) Received: (from toasty@localhost) by home.dragondata.com (8.9.2/8.9.2) id MAA09057; Mon, 7 Jun 1999 12:04:46 -0500 (CDT) From: Kevin Day Message-Id: <199906071704.MAA09057@home.dragondata.com> Subject: Re: Full filesystem In-Reply-To: <375BFAAE.CCC65835@nall.com> from Joe Nall at "Jun 7, 1999 12: 0:30 pm" To: joe@nall.com (Joe Nall) Date: Mon, 7 Jun 1999 12:04:46 -0500 (CDT) Cc: freebsd-isp@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL43 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-isp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > spork wrote: > > > > Hi, > > > > We're running a news server that does a "rolling expire", so it is capable > > of keeping the drives 99% full (31GB) 24/7. We have two 34G ccd arrays, > > so even bumping down one percent is kind of a waste of space. Since newfs > > gives a buffer zone for safety (you can fill the fs to 105% or so), I want > > to keep usage up above the warning zone. > I'm not an expert on this filesystem, but on other filesystems that > extra 5-10% is used by the filesystem to minimize fragmentation. On > HP-UX there is a definite performance impact when this is made too > small. Can a filesystem guru clue us in on the right rule of thumb? Is > it 10% or is there a particular total size (e.g. 500mb) above which the > extra space is wasted? I'm going to do some big raids this summer and 5% > of 100+GB starts to add up. > To quote 'tunefs' -m minfree This value specifies the percentage of space held back from nor- mal users; the minimum free space threshold. The default value used is 8%. This value can be set to zero, however up to a fac- tor of three in throughput will be lost over the performance ob- tained at a 10% threshold. Settings of 5% and less force space optimization to always be used which will greatly increase the overhead for file writes. Note that if the value is raised above the current usage level, users will be unable to allocate files until enough files have been deleted to get under the higher threshold. So, keep that in mind. :) Kevin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" in the body of the message