Subject: Re: [linux-audio-dev] Performance and SCSI
From: Karl JH Millar (kmillar_AT_MIT.EDU)
Date: Sat Apr 22 2000 - 20:30:35 EEST
Paul,
I was trying to understand why the performance you get is so much better than
what Benno and I get, and I'm now fairly convinced it's the hardware. Your
SCSI adapters can reorder IO in order to minimize seeking, so the
metadata updates when using O_SYNC don't kill performance as they do in
the case of IDE drives. I wouldn't be surprised to find out that it delays
writing if a read has been requested either, so you don't get big delays when
bdflush runs.
I'm sure that the low latencies and large buffers on your drives don't hurt
any either :)
The upshot of this is that for a SCSI drive, good performance is easy to get
irrespective of how you do your I/O. For IDE drives, the kernel and app need
to work together to try to do what the adapter is giving you automatically.
Karl.
This archive was generated by hypermail 2b28 : Sat Apr 22 2000 - 21:08:49 EEST