Re: Timestamp vs Date/time Decimal fields, Overhead in RPG programs...

On 18/03/2010, at 11:37 PM, Charles Wilt wrote:

I'm a little surprise to see RPG III program calls performing better
than RPG IV program calls..

I'm not. ILE dynamic program calls have more work to do. I mentioned earlier that much effort was expended on improving the ILE dynamic call path.

In ILE the focus should be on using bound procedure calls instead because they are so much faster than any form of dynamic program call. However, that doesn't mean ILE dynamic calls should not be used when appropriate. As always the primary driver should be application maintenance and modularity rather than purely performance.

I thought perhaps it was activation group related. Initially, I had *NEW
calling *CALLER. I tried changing both program to run in the same named
activation group. That had no effect. So I tried changing both to
DFTACTGRP(*YES). That had a little effect, lowering the times to 11s and 110s.

I'd expect no difference between *NEW and *CALLER versus same named activation group. In both case all the work involved in setting up for the call will happen in the caller's activation group thus avoiding the creation of another activation group. Not sure why using the default activation group should be different except perhaps it takes slightly less time to locate it?

Scott's point about the extra call stack entry is on the money. The ILE call path is at least twice as long as the OPM one (plus any activation group creation hit)..

Regards,
Simon Coulter.
--------------------------------------------------------------------
FlyByNight Software OS/400, i5/OS Technical Specialists

http://www.flybynight.com.au/
Phone: +61 2 6657 8251 Mobile: +61 0411 091 400 /"\
Fax: +61 2 6657 8251 \ /
X
ASCII Ribbon campaign against HTML E-Mail / \
--------------------------------------------------------------------