Date: Sat, 12 Aug 1995 02:39:06 -0700 (PDT) From: "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com> To: jmb@kryten.Atinc.COM (Jonathan M. Bresler) Cc: msmith@atrad.adelaide.edu.au, john@zyqad.co.uk, freebsd-hardware@freebsd.org Subject: Re: Upgrade to my machine Message-ID: <199508120939.CAA05542@gndrsh.aac.dev.com> In-Reply-To: <Pine.3.89.9508111506.I6515-0100000@kryten.atinc.com> from "Jonathan M. Bresler" at Aug 11, 95 03:50:40 pm
next in thread | previous in thread | raw e-mail | index | archive | help
... > > > IOZONE: auto-test mode > > > > Pretty typical numbers, but the numbers I use are always for a > > freshly newfs disk using the outer tracks (ie, maximal conditions). > > > > Your numbers are considerably lower than my 2.4/4.4MB/sec rating > > for the DSP3053L due to file system fragmentation and the fact your > > probably running at about the 250MB mark in on the disk (guessed from > > your numbers :-)). > > at that point the disk had been in constant use for some 5 > months, must have had 300mb on it at the time Okay, so I was off 50MB, that's within 10%. Don't I get 2 points for that?? :-) :-) > > > bonnie: > > > > > > /* dont jive with the iozone number very well, do they ? comments ? */ > > > > Was this under identical conditions? > > identical is a pretty strong condition.....the fujitsu drive was > new at the time so fragmentation was minimal. at this time, several > weeks of use it still reports the same bonnie numbers. Is what I really meant was Iozone and Bonnie run one right after the other so that they system had not changed as far as fragmentation, where on the disk the test where run, etc. > > > disklabels: NOTE: any misconfiguration here is my fault > > > any recommenations are appreciated > > > > Actually, any miss configuration on the DSP3053L is _MY_ fault. That > > is my standard label, I recognize it! > two points! :-) > > > type: SCSI > > > disk: sd1s1 > > > > This one you can take the blame for, but it is reasonable. > > ;) had a place to take clues from. :-) > > > > /* sysinstall put all these partitions off cylinder boundaries.....hmmm */ > > > > You don't have to worry about cyl boundaries inside the BSD slice. > > excellent....but ths slice must be on cyl boundaries ?? To maintain compatibility with what almost all other OS's do, YES. Strickly speaking for FreeBSD itself, NO it could care less about that. Though some BIOS's might have a problem with a bootable partition that did not start on a track boundary. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199508120939.CAA05542>