RE: SORT on the iSeries...

["Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>]

> From: Michael Rosinger
>
> List,
>
> This is an iSeries newbie question so please bare with me. We are getting
> started on the project of migrating applications (all batch) from VSE
> (mainframe) to iSeries. I was "told" by someone that is a long-time AS/400
> person that "there is no system sort on the iSeries". I find that very
> hard
> to believe. There must be some system provided sorting mechanism. I see
> that
> ILE/COBOL fully supports the SORT verb so that must be interfacing with a
> system-provided utility.

Michael already told you about FMTDTA, which is the iSeries version of SORT.
It uses the same external sort specifications that I've used for 30 years,
which I think are compatible with the mainframe.

However, this is another area where you might want to consider taking
advantage of certain iSeries features.  In this case, you might want to
create a logical file over the physical file.  One of the options you have
when creating a logical file is to set it up as MAINT(*DELAY), which means
that the index is not kept up to date with every record addition so there's
no additional overhead to programs that maintain the file, but when a
program opens the logical file (which should only be the program that
actually uses that logical), the index is rebuilt.  In effect, the data is
automatically sorted when you need it.

Why do this?  Because if you ever need to run the same process twice in a
row, for example, then the second run wouldn't need to rebuild the access
path; it would already by in place.  And it seems to me that creating an
access path in general is a lot faster than physically resequencing the data
in a file.

I don't want to say you'd always run faster; I'm not certain that rebuilding
the access path for just a few new records is any faster than rebuilding it
for the whole file; someone with more knowledge would have to say.  And
using an index to access records from a file is not the same as reading
records sequentially in blocks.  If you sort a large file once and run many
reports over that file, FMTDTA may still be the way to go.

But recognize that you have options.  (Another option is OPNQRYF, which we
can leave for another day <grin>.)
 
Joe