From owner-freebsd-questions Fri Apr 17 15:54:38 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA25860 for freebsd-questions-outgoing; Fri, 17 Apr 1998 15:54:38 -0700 (PDT) (envelope-from owner-freebsd-questions@FreeBSD.ORG) Received: from caladan.tdx.co.uk (caladan.tdx.co.uk [195.188.177.4]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA25803 for ; Fri, 17 Apr 1998 22:54:24 GMT (envelope-from kpielorz@tdx.co.uk) Received: from tdx.co.uk (lorca-tx.tdx.co.uk [195.188.177.242]) by caladan.tdx.co.uk (8.8.8/8.8.8) with ESMTP id XAA00746; Fri, 17 Apr 1998 23:54:05 +0100 (BST) (envelope-from kpielorz@tdx.co.uk) Message-ID: <3537DD6D.73132BD7@tdx.co.uk> Date: Fri, 17 Apr 1998 23:53:33 +0100 From: Karl Pielorz Organization: TDX X-Mailer: Mozilla 4.04 [en] (WinNT; I) MIME-Version: 1.0 To: Vallo Kallaste CC: freebsd-questions@FreeBSD.ORG Subject: Re: One more ccdconfig question References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Vallo Kallaste wrote: > > On Fri, 17 Apr 1998, Karl Pielorz wrote: > > > In truth it is _very_ hard to work out the 'best' interleave factor for > > CCD... > *** > > Ok, this is very nice answer, too ;) Thanks ;-) > Very informative, but I know all this, so it can't help. I think this > needs very good knowledge about disks and controllers and some > programming experience in this area, huh. Unfortunately I don't have > any. Nope, I think at the end of the day - because of all the variables involved you need _LUCK_ and lots of it :-) You could spend _ages_ working this out with different loads, different files sizes, different read to write ratios etc. - but still not arrive at the 'mathematically' correct solution :-( > Btw, what program or method you suggest for benchmarking ? I mean > something which simulates _real_ workload and puts hard stress for disk > subsystem. All this must be measurable and comparable, certainly. Hmmm... I've used everything from make-worlds to writing out huge tar's up into the gigs region, not very scientific - and at the end of the day the figures only made me 'feel' happy, I'm sure they wouldn't stand up in court... > Thanks No problem... If I were you, I'd 'pick & play' - i.e. just go for a number... Try 4096, it's a nice number... Sorry I can't be more 'scientific', but I've found theres just too many different things involved... :-) Regards, Karl To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message