From owner-freebsd-hackers Sun Jun 18 11:57:03 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id LAA26872 for hackers-outgoing; Sun, 18 Jun 1995 11:57:03 -0700 Received: from haywire.DIALix.COM (peter@haywire.DIALix.COM [192.203.228.65]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id LAA26862 for ; Sun, 18 Jun 1995 11:56:59 -0700 Received: (from peter@localhost) by haywire.DIALix.COM (8.6.12/8.6.12/DIALix) id CAA09843; Mon, 19 Jun 1995 02:56:42 +0800 Date: Mon, 19 Jun 1995 02:56:40 +0800 (WST) From: Peter Wemm To: Ollivier Robert cc: hackers@freebsd.org Subject: Re: How does the disk IO clustering work? In-Reply-To: <199506181834.UAA14935@blaise.ibp.fr> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: hackers-owner@freebsd.org Precedence: bulk On Sun, 18 Jun 1995, Ollivier Robert wrote: > > > > > jhome # tunefs -p /dev/rsd0h > > tunefs: maximum contiguous block count: (-a) 1 > > You can push this to 8 > > > tunefs: rotational delay between contiguous blocks: (-d) 4 ms > > You should lower this to 0. Thanks! > > tunefs: maximum blocks per file in a cylinder group: (-e) 1024 > > tunefs: minimum percentage of free space: (-m) 10% > > Now the default is 8%. Just as a btw, I remember it being 5% for a while. When I looked at the code, I thought I read that if the minfree was <= 5%, then the auto switch from space to time was disabled. I know there was some sort of hysteresis(sp?) problem as well. I've seen both of these on SVR4 systems - if you set minfree to 5%, it will automatically shift from time to space, but will never shift out automatically. At 6% it shifted out Ok, but it oscilates too quickly. > There is another parameter that can be set only at newfs > time : -n 1 Damn! Drat! Thanks, -Peter > -- > Ollivier ROBERT -=- The daemon is FREE! -=- roberto@FreeBSD.ORG > FreeBSD keltia 2.0-BUILT-19950503 #3: Wed May 3 19:53:04 MET DST 1995 >