Date: Mon, 26 Aug 1996 08:32:59 -0700 From: Darryl Okahata <darrylo@hpnmhjw.sr.hp.com> To: current@freefall.freebsd.org Cc: "Rodney W. Grimes" <rgrimes@GndRsh.aac.dev.com> Subject: Re: Speedingup the "worldstone" Message-ID: <199608261532.AA141643579@hpnmhjw.sr.hp.com> In-Reply-To: Your message of "Sun, 25 Aug 1996 23:08:11 PDT."
next in thread | raw e-mail | index | archive | help
Rodney W. Grimes <rgrimes@GndRsh.aac.dev.com> wrote:
> More correctly you won't get more than 6 to 8MB/s on ISA based 486
> systems,
On an ISA-based system, it's a lot closer to 2MB/s (sustained, not
burst). (Or, are you talking about RAID/striping?)
Here are some bonnie numbers for a Quantum Fireball 1280S. Note
the huge improvement going from ISA to PCI. The first is for an
ISA-based 1542CF controller, and the second is for an NCR815/PCI-based
controller:
-------Sequential Output-------- ---Sequential Input-- --Random--
-Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU
ISA/1280 60 1349 95.2 2237 63.9 804 35.7 1361 92.2 1943 22.0 42.6 5.7
PCI/1280 256 4267 86.3 4393 19.2 1207 8.3 4118 77.9 4620 17.8 56.8 2.5
The exact same drive mechanism was used on two systems:
* 486DX4/100 w/256K cache, 24MB RAM, & an ISA-based Adaptec
1542CF. FreeBSD 2.1R, untuned kernel, untuned 1542 (DMA setting
of 5MB/sec).
* P133 w/512K cache, 64MB RAM, & an NCR815-based PCI SCSI
controller. FreeBSD 2.1R, untuned kernel
The block numbers above should be CPU-independent, although I'm not sure
about the "per-char" ones; these might be impacted by CPU performance.
-- Darryl Okahata
Internet: darrylo@sr.hp.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Hewlett-Packard, or of the
little green men that have been following him all day.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199608261532.AA141643579>