Date: Fri, 10 Feb 1995 11:55:44 -0800 (PST) From: julian@tfs.com (Julian Elischer) To: hasty@netcom.com (Amancio Hasty Jr) Cc: FreeBSD-hackers@wcarchive.cdrom.com Subject: Re: My iozone results Message-ID: <m0rd1RQ-0003wgC@TFS.COM> In-Reply-To: <199502100637.WAA18562@netcom14.netcom.com> from "Amancio Hasty Jr" at Feb 9, 95 10:37:51 pm
next in thread | previous in thread | raw e-mail | index | archive | help
I gather from this that the 2842 is 'mostly' reliable :) from the talak we had with Kirk McKusick the other day.. try the following.... set the disk up with a disklabel that suggests a geometry of: 1 head, 4096 sectors per track.. and then repeat the test.. (or newfs the partition with that geometry if it's possible) and let us see if there is a difference... julian > > > Anyone got better iozone results... > > MB reclen bytes/sec written bytes/sec read > 1 512 1890390 3195660 > 1 1024 3121342 4194304 > 1 2048 3627506 4628197 > 1 4096 3728270 4793490 > 1 8192 4067203 4793490 > 2 512 2294320 3234162 > 2 1024 3234162 4194304 > 2 2048 4067203 4549753 > 2 4096 4260880 4709393 > 2 8192 4129776 4628197 > 4 512 2396745 3314017 > 4 1024 3376546 4473924 > 4 2048 4294967 4925421 > 4 4096 4588640 4925421 > 4 8192 4588640 4925421 > 8 512 2407492 2760261 > 8 1024 3452546 4569114 > 8 2048 4400581 5041041 > 8 4096 4858560 5064819 > 8 8192 4858560 5064819 > > ----- > My system is a P66 VL with a Quatum Empire 2100 and a Adaptec 2842a VL. > > Tnks, > Amancio >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?m0rd1RQ-0003wgC>