[Ti] [Ti]Journaling, was --> Journeling.

[Kynan Shook]

Actually, most drives write faster than they read; I'm not sure exactly 
why this is the case, but here are several "drive comparison" pages 
from barefeats.com that can back me up; the only tests that aren't 
totally behind my position here are the 15,000 RPM Ultra320 SCSI 
drives, but they're really all over the map; some support you, some 
support me.  The other two pages (serial ATA and notebook drives) show 
that writing is generally faster than reading.  I do, however, 
completely agree with your random versus sequential analysis; it takes 
the drive a little bit of time to jump around and find the right 
cylinder to read from.
http://www.barefeats.com/hard27.html
http://www.barefeats.com/hard31.html
http://www.barefeats.com/hard28b.html

On Aug 7, 2004, at 12:22 PM, b wrote:

> A few 'pointers' drives read faster thn they write, and they read 
> sequentially-oriented files faster than randomly located files.
>
> Packet size, state of the drives fragmentation and free space, and 
> size of the file also affects speed (not just the 'time', but actual 
> per second transfer rate).