There are a few batch jobs which have a high elapsed time. I had run a STROBE on the job and did some analysis.It seemed that the %CPU for the statements which executed code, while executing code in a version of natural different from the version of the program was more, than when the code executed was in the same version.
More specifically, one case was when, the program was in NAT 3.1 but it was making use of GDA and PDA in NAT 2.2. The STROBE highlighted the statements in the DEFINE DATA for the GDA and PDA to be taking more CPU.
My question is, how far does executing different versions of natural influence the performance. In the case specified above, will I find better performance by stowing all the data areas in the same version as the program.
Lastly, from the STROBE analysis, it pointed out the below procedure to consume high CPU:
MODULE NAME - .NATNUC SECTION NAME- NAT316SH PROCEDURE NAME - NATFLIP NATURAL SYSTEM SERVICES
What does this NATFLIP procedure stands for? What does it do?