RE: opnqry efficiency question

Mark,

Query engine performance is one of my little hobbies.  In my experience, if
the amount of data involved is significant, a PROPERLY TUNED opnqryf or SQL
statement will perform as well as or better than an RPG program making the
same selection.  Note the emphasis on PROPERLY TUNED - if you just take the
defaults, don't look at the tuning recommendations available under debug,
and don't take your database structure and content into account when
formulating your record selection and ordering criteria, then you're playing
a crap shoot - maybe you win, maybe you don't.  One biggie - look at the
help text for the ALWCPY parameter, and use ALWCPY(*OPTIMIZE) when it fits,
so that it will use the sort engine instead of the query engine when it's
appropriate.  There's TREMENDOUS power available with that.

Also, if you don't make the complementary changes to your RPG to use the
OPNQRYF output efficiently, you're artificially constraining the performance
from the back end.  For example, if you're just reading the OPNQRYF output
sequentially, lose the 'K' in the F spec - you don't need it and it slows
you down.  Obviously, having your record selections in BOTH an OPNQRYF and
RPG will slow you down as well.  And if OPNQRYF is selecting the records,
SETLL/READE processing is going to really slow it down.

Dave Shaw
Spartan International, Inc.
Spartanburg, SC
---
If you would like to subscribe to the MAPICS-L mailing list send email to
MAPICS-L-SUB@midrange.com or go to www.midrange.com and follow the
instructions.

-----Original Message-----
From: Mark A. Manske [mailto:mmanske@minter-weisman.com]

Hi all;

I am not sure if I am from the old school or the new - here is my question;
I have proved in one particular program my way is more efficient; however I
am wondering
which way is "truly" more efficient as my IS manager was "taught" one way,
and I forget now if I was taught or just out of being a "control freak" in
RPG and not
wanting to relinquish anything to C/L that I code the way I do...

He has always tried to do opnqry before the program call to trim down the
data as far as he can before calling the program - I do not have faith in
opnqry as I have seen
it drop records it should not have (PTF's "probably fixed this) so I tend to
just take the file as is and code for my selection criteria.

In a test today running it his way took 39 seconds, my way took 17 seconds.

Now I do not know if this would hold true in all instances..

I guess I am wondering at a bare metal level, say you leave the key as is,
but only put selection criteria in, does it matter...
On the flip side, just how much system resources are "wasted" say for
creating a opnqry over a file with a completely different key??

Any thoughts??

TIA

Mark A. Manske
[mailto:mmanske@minter-weisman.com]
Sr. Project Lead
Minter-Weisman
+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---