Re: bored and setting up a 3/4tb array at the office.
On Wed, 9 Apr 2003, Lennart Sorensen wrote:
> On Tue, Apr 08, 2003 at 07:31:16PM -0400, Mike Dresser wrote:
> >
> > Version 1.03 ------Sequential Output------ --Sequential Input- --Random-
> > -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
> > Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP
> > mike 1G 25926 89 113641 36 29163 7 30287 96 49783 4 507.5 0
> > ------Sequential Create------ --------Random Create--------
> > -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
> > files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
> > 16 2830 91 +++++ +++ +++++ +++ 3093 99 +++++ +++ 9060 100
> > mike,1G,25926,89,113641,36,29163,7,30287,96,49783,4,507.5,0,16,2830,91,+++++,+++,+++++,+++,3093,99,+++++,+++,9060,100
> >
> > /dev/sda:
> > Timing buffer-cache reads: 128 MB in 0.34 seconds =376.47 MB/sec
>
> Yep a P4 has good memory IO.
>
> > Timing buffered disk reads: 64 MB in 1.43 seconds = 44.76 MB/sec
>
> I have seen better speeds on single disks than that. That card should
> be able to pull off much better numbers than that, although perhaps with
> those slow poke drives it can't.
Much better values under raid5, somewhere in the 70's on hdparm.
Here's the Raid5 values:
Version 1.03 ------Sequential Output------ --Sequential Input- --Random-
-Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP
mike 1G 18574 64 20326 6 9319 2 28163 89 94607 8 381.1 0
------Sequential Create------ --------Random Create--------
-Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
16 2643 84 +++++ +++ +++++ +++ 3178 98 +++++ +++ 9712 90
mike,1G,18574,64,20326,6,9319,2,28163,89,94607,8,381.1,0,16,2643,84,+++++,+++,+++++,+++,3178,98,+++++,+++,9712,90
Basically goes from a super fast writer, to a super fast reader. Kind of
expected that, actually.
Mike
Reply to: