Re: Paging file

[vhamberg@xxxxxxxxx]

James

Say you have a 6meg file. On a *nix swap partition it'll
be a continuous block of 6meg, right? On a 400, it'll
probably be in 6 1meg extents, on 6 separate drives. All
6 will go to their extent at basically the same time, so
seek time is about the same, since it's in parallel. the
actual elapsed time to physically read the data off
disk, assuming the same rotational speed, is about 1/6th
that of the long, 6meg file on a single platter.

Eh?

Vern
> On Tue, 29 Oct 2002, Dennis Lovelady wrote:
>
> > How good is a swap file's use of disk heads?  If you have, say, forty disks
> > on your system, and you create a swap file, are you careful to allocate
> > equal parts from each of your disk drives, or are you going to busy one or
> > two of the disks with the swap space?  iSeries uses (read "takes advantage
> > of") all heads / arms on the system, providing a far better method of
> > storage (sans human effort) than any other system I've seen.
>
> I believe (i.e. I don't know) that in this case a dedicated swap partition
> should actually outperform the "spread out over all the disks" case
> because the swap data is "compact", meaning that it is contiguous.  One
> disk arm reading a continuous block of data should beat many arms seeking
> all over several disks.
>
> But without benchmarks, tests, case studies for specific applications,
> etc. we will probably never know.
>
> Of course arguments to the contrary are welcome and read with interest.
>
> James Rich