Date: Tue, 26 Aug 1997 10:26:06 -0700 (PDT) From: Simon Shapiro <Shimon@i-Connect.Net> To: dg@root.com Cc: current@FreeBSD.ORG Subject: Re: IDE vs SCSI was: flags 80ff works (like anybody doubted it) Message-ID: <XFMail.970826102606.Shimon@i-Connect.Net> In-Reply-To: <199708261514.IAA12291@implode.root.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi David Greenman; On 26-Aug-97 you wrote: > >FreeBSD does ALL its file I/O in 4K blocks. Your test is really a test > >of the kernel's ability to issue strategy calls quickly, at best. > > Huh? No it doesn't; what ever gave you that idea? > > -DG Which idea? The 4K? I enable a certain debugging level (DPT_DEBUG_COMPLETION) in the DPT driver, lean back and enjoy the messages fly by. I have not seen anything larger than 4K yet. In SCSI, random disk I/O, up to about 8K blocks, the amount of time it takes to actually process the transaction is about the same, regardless of block size. We tested it on SPARC20 with Solaris 2.5.1, Linux with DPT, Linux with AHC2940, FreeBSD with AHC2940, and FreeBSD with DPT. Again, we did not measure megabytes per second, but operations/sec. Simon
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.970826102606.Shimon>