RE: Join 2 files with 1 key from each file.

Thanks Matt.

> ------------------------------
> 
> message: 2
> date: Wed, 4 May 2005 12:02:15 -0600
> from: "Tyler, Matt" <mattt@xxxxxxxxxxxxxx>
> subject: RE: Join 2 files with 1 key from each file.
> 
> John,
> 
> You cannot join and index two files with fields from each file in the
> index.
> A join LF will not allow and field from the secondary file(s) in the
index
> and you cannot have an index in a SQL VIEW.  It depends on how you are
> planning on accessing the data.
> 
> You need to start debug and then look at the job log message after the
SQL
> is finished to see what it did not make the join work.  I have had to
make
> programs that respond poorly for the sole purpose of pleasing the end
> user.
> 
> 
> Your SQL example below is never going to give good results in large
data
> sets.   The system will more than likely have to build a temporary
data
> set
> so that it can complete the data ording.  Check out the debug messages
and
> see if that is the case.
> 
> You want to order by fields not in the join.  Is there a way to use
those
> fields in the join as well?  Is there an index that starts with those
two
> fields?

Nope, can't use those flds in the join as well.

There's index(s) but they don't seem to be helping much.

The optimizer is finding indexes, I'm thinking the file has too many
records.  There's a redundant sentence <g>.

> 
> 
> Thank you,
> Matt Tyler
> WinCo Foods, LLC
> mattt@xxxxxxxxxxxxxx
> 
> 
> -----Original Message-----
> From: rpg400-l-bounces@xxxxxxxxxxxx
[mailto:rpg400-l-bounces@xxxxxxxxxxxx]
> On Behalf Of Rusling, John B. (Alliance)
> Sent: Wednesday, May 04, 2005 11:22 AM
> To: rpg400-l@xxxxxxxxxxxx
> Subject: Join 2 files with 1 key from each file.
> 
> I have 2 files I want joined and indexed on 1 key from each of the 2
> files.
> 
> The 2 files are named s1 and s3, both have the common field of Order#
> (on) in them
> 
> s1 JOIN s3 ON s1on = s3on .
> 
> I want the joined file indexed by s1name, then s3itno
> 
> ORDER BY s1name, s3itno .
> 
> 
> I'm using embedded sql in rpg and everything works fine except the
> response
> time on one of our production machines gets to be 5+ minutes.!
> 
> ---->
> I want to have a prebuilt, index or logical so that when this program
is
> used, the response time will fall into the acceptable range.
> 
> I've built most every combination of logical file (using dds) to this
> end
> but the response time is still horrid.
> 
> The s1 file has 2.25 million records and the s3 file has about 2.8
> million.
> 
> Can anyone give me some guidance on this please?  I've tried
everything
> I can think of for the last day or 2 and response time is still
crappy.
> 
> 
> Thanks for any help at all,
> 
> John B.
> 
> --
> 
> This is the RPG programming on the AS400 / iSeries (RPG400-L) mailing
list
> To post a message email: RPG400-L@xxxxxxxxxxxx
> To subscribe, unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/rpg400-l
> or email: RPG400-L-request@xxxxxxxxxxxx
> Before posting, please take a moment to review the archives
> at http://archive.midrange.com/rpg400-l.
> 
>