Date: Fri, 2 Jun 1995 01:07:47 -0700 (PDT) From: "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com> To: msmith@atrad.adelaide.edu.au (Michael Smith) Cc: FreeBSD-hackers@FreeBSD.ORG (FreeBSD hackers) Subject: Re: A performance mystery Message-ID: <199506020807.BAA06938@gndrsh.aac.dev.com> In-Reply-To: <199506020758.RAA15903@genesis.atrad.adelaide.edu.au> from "Michael Smith" at Jun 2, 95 05:28:08 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> > > > Which 'oldish' disk? Are we talking, say, ST11200's? > > > > The disk is a SEAGATE ST43400N > > Hmm, rules out the smarts in the drive; the Elite 3's are pretty quick, > but I wouldn't have thought that'd make the difference. > Elite 3's are quick to read but horrible slow at writes, they do *not* have a very good write behind cache in them. I was typically seeing 3.5MB/sec read but only 800KB/sec writes to these disk on wcarchive (a PILE of them). This was under FreeBSD 1.X so the cluster code may have helped this. Try running ``iozone 128 8192'' to make sure you get well outside of the buffer cache. I remeber you had a lot of memory in this box so your iozone read results where most likely pure cache unless you ran something other than auto. Better yet run it on the raw disk to see what the drive can really deliver. > > > And which connor? > > > > Dont recall, but the Micron is all of a week old, I remember looking for the > > specs for the drive and couldnt find it, too new. > > Not even bundled with it? Sheesh. Most OEM's buy there drives in OEM quantities, that means they get 1 copy of the OEM full blown manual (usually an inch or two thick) and 1000's of drives :-). They don't have anything to pass on to the end user. > > > What sort of filesystem layouts do you have? Is the compaq working on just > > > one drive, or several? > > > > Both on one drive, Compaq has a 2500mb / 300mb swap, the Micron a 800M / and > > 200M swap. > > You're saying you have a single 2.5G filesystem with everything on it? > Same blocking factors as the 800M filesystem? > > Fresh out of ideas, I'm afraid. I think the disk performance reported is not reflective enough of the drive to really know. Where both drives freshly newfs'ed with the exact same parameters. If not that can be a big difference. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Custom computers for FreeBSD
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199506020807.BAA06938>