Date: Sat, 6 Nov 1999 05:51:08 -0800 From: John-Mark Gurney <gurney_j@efn.org> To: David Miller <dmiller@search.sparks.net> Cc: Rob Snow <rsnow@lgc.com>, freebsd-questions@freebsd.org Subject: Re: writing much slower than reading... Message-ID: <19991106055108.38507@hydrogen.fircrest.net> In-Reply-To: <Pine.BSF.3.96.991106074121.8537A-100000@search.sparks.net>; from David Miller on Sat, Nov 06, 1999 at 07:45:16AM -0500 References: <38240CC0.8099D19D@lgc.com> <Pine.BSF.3.96.991106074121.8537A-100000@search.sparks.net>
next in thread | previous in thread | raw e-mail | index | archive | help
David Miller scribbled this message on Nov 6: > John is commenting on the 1.68 sec cpu time required to write 128 MB. > > Isn't this just a normal result of IDE vs scsi? If IDE is going to be as > fast and cpu un-intensive as scsi I wasted a lot of money on my last bunch > of servers:) IDE has improved a bunch w/ DMA support and other things... sure, it still has limitations, but for a single drive system, IDE is great... you are problably thinking back when IDE didn't support busmastering, and required poll io to get the data across... and actually, IDE has a lower command latency than SCSI (see Bruce Evans's posting about it on -hackers or -current a long time ago)... IDE's Problems: two drives can't talk on the channel at once until the other has completed can't queue up commands like scsi only two drives per channel SCSI's Problems: expensive wiring can be tricky because of supporting 7 or 15 devices -- John-Mark Gurney Voice: +1 408 975 9651 Cu Networking "The soul contains in itself the event that shall presently befall it. The event is only the actualizing of its thought." -- Ralph Waldo Emerson To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19991106055108.38507>