RE: Performance question of RPG native read vs SQL

I agree with what Scott wrote.

Also, try also to focus on record selection in the RPG program. I worked
with a program once that read a HUGE G/L file row by row. It was reading
billions of rows and turns out it only selected less than 10% of those.
When SQL was used, the selection was alot faster since only a small amount
of data was actually selected for processing instead of the whole file.
I'd say that the record selection may be what the culprit in your case.
Personally, I would not worry about how many bytes are actually used. I
would get SQL to pull in what you need and do the summary too. Just make
sure the indexes are adequate. At least have them on what you are
selecting by. If you can do that, and read the SQL cursor fetching 400+
rows at a time, you'll most likely see improvement.

Another thing to look at is a driver file, if you have it. The program I
referenced above was driven by another file whose contents determined what
to select, but the program slogged through the whole file each time.
Linking the driver to the selection also sped it up too. I don't remember
the specifics, but the improvement in runtime was definitely noticeable by
IT and the users.

Hope this helps,

Doug


"CONFIDENTIALITY NOTICE: This e-mail transmission (and/or the attachments accompanying it) contain confidential information belonging to the sender. The information is intended only for the use of the intended recipient. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution or the taking of any action in reliance on the contents of the information is strictly prohibited. Any unauthorized interception of this transmission is illegal under the law. If you have received this transmission in error, please promptly notify the sender by reply e-mail, and then destroy all copies of the transmission."