Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!news.wildstar.net!news.ececs.uc.edu!news.kei.com!newsfeed.internetmci.com!howland.erols.net!EU.net!Ireland.EU.net!maths.tcd.ie!not-for-mail From: dwmalone@maths.tcd.ie (David Malone) Newsgroups: comp.unix.bsd.freebsd.misc Subject: Re: ccd driver Date: 26 Sep 1996 13:09:20 +0100 Organization: Dept. of Maths, Trinity College, Dublin, Ireland. Lines: 30 Message-ID: <52drpg$oo1@hamilton.maths.tcd.ie> References: <52ckjv$k5j@itchy.serv.net> NNTP-Posting-Host: hamilton.maths.tcd.ie zeno@serv.net (Sean T. Lamont) writes: >Can someone comment about an interleave factor? The disks I was using had >different geometry and I was using 16 as an interleave. In reading the >documentation it's not clear as to what this should be. Here are some measurements I did with two 2GB connor file pro's a few months agao - they were connected with Adaptec 2940's I think. The measurements were done with iozone I think. I should have used bonnie - but the disk is in use as a news spool. I've seen 1 cyl recomended for news spools, so I think this is what I setteled on. I don't have figures for the original disks so.. David. Same Controler Different Controller Interleave W/R Speed (MB/s) W/R Speed (MB/s) 8 1 16 1.6 5.4 32 2.1 5.8 64 2.9 6.7 128 4.5 7.3 5.4 7.5 256 5.3 7.2 5.4 7.6 512 5.9 7.1 5.9 7.2 1024 6.1 6.8 6.1 6.8 (1cy) 1044 6.3 6.7 6.3 7.0 2048 6.4 6.5 6.5 6.5 (2cy) 2088 6.2 6.5 6.2 6.6 4096 5.7 6.4 5.8 6.4