Re: KEYFILE parm on RGZPFM

Benchmarking this would be very interesting.  The physical reorg of the
records is undoubtedly slower, since they have to be read in key order.
However, the rebuild of that particular index should be faster.  I think the
net is probably always "somewhat" slower, but I don't know how big
"somewhat" is, or how it varies with the complexity of the index(es ) over
the file.  I don't have a machine I can benchmark on right at the moment,
though.

Dave Shaw
Simpsonville, SC
---
If you would like to participate in the MAPICS-L mailing list go to
http://lists.midrange.com/cgi-bin/listinfo/mapics-l and follow the
instructions.

----- Original Message -----
From: "Dan Rasch" <drasch@mail.win.org>
>
> I just learned that specfying KEYFILE(*FILE) on a RGZPFM will reorganize
> the file into the keyed order.  Sounds great for performance, but I am
> wondering how it affects the reorg itself.  Does it add alot/a little time
> to the reorg, or nothing appreciable?
>
> Dan Rasch - because if the human species concentrated on the really
> important things in life, there would be a shortage of fishing poles!
> IBM Certified twice........ but never noticed.