Actually the physical disk position can make a HUGE difference. I have
a machine with a velociraptor 1GB drive which I have parititioned into two
500GB partitions, the average speed of the first (on the outside cylinders
of the drive) is far better than the second on the inside. Sorry I don't have
specific numbers now but the machine is in service so I can't bash on it
anymore. However, it was a difference of almost 50%.
Modern drives don't have a fixed number of sectors/track. Instead they
try to keep a fixed density so there are more cylinders on the outer tracks
than the inner. Thus per revolution more sectors go past the head for the
tracks on the outside (low numbered tracks) verses those on the inside.
Given that modern drives have interface speeds that are considerably
faster than the actual media I/O speed, the media I/O rate is the limiting
factor and thus where on the disk your data lies makes a big difference in I/O
throughput.
For those reasons I always try to put swap and I/O critical stuff, like
swap, at the beginning of the drive and loath partitioning software that thinks
it's smarter and puts things where it wants.
>The variation in performance indicates a problem of variable
>severity like fragmentation or the position on the physical disk - but I
>don't think either of those are likely causes, because there's only one
>file in each volume, and physical disk position shouldn't have such a
>marked effect should it? Any other suggestions?
>Thanks,
>Julian
_______________________________________________
CentOS-virt mailing list
CentOS-virt at centos.org
http://lists.centos.org/mailman/listinfo/centos-virt