Re: ILE V OPM Performance Increase?

[Jerry Adams <jerry@xxxxxxxxxxxxxxx>]

Doc,


Quantitatively (benchmark-wise) I can't say which (OPM or ILE) is 
faster.  But your last post said anecdotal as well, so here goes.


Our base applications are still running in the /36 environment with RPG 
II programs and procedures.  However, every new function/program is 
either RPG IV or ILE RPG.  I recently, for example, had to add a new tax 
structure (ILE) to our existing on-line pricing module (RPG II).  The 
ILE program had (currently) fifteen (15) subprocedures.  The response is 
sub-second.


As for batch, I have no quantitative figures here either.  But I quit 
using the SBMJOB simply because on our i/5 520 (and even on the old 720) 
it takes seconds to process files with 100k records. 


The 520 probably forgives a lot of any inefficiencies I may 
(inadvertently) foist upon it.  And I dare say that, if I wanted to do 
so, I could bring an ILE or IV program to its knees.  We don't have any 
RPG III programs, but I have noticed no degradation in on-line response 
times when putting hooks (CALLs) from an RPG II program to an ILE or IV 
program.  Plus the fact that I can easily add new tax structures to the 
above mentioned program and a myriad of other functions, such as IFS 
access, that simply cannot be done (or not done easily) in OPM makes 
eschewing OPM a no-brainer.


        * Jerry C. Adams
*iSeries Programmer/Analyst
B&W Wholesale Distributors, Inc.* *
voice
        615.893.8633x152
fax
        615.995.1201
email
        jerry@xxxxxxxxxxxxxxx <mailto:jerry@xxxxxxxxxxxxxxx>



e.h.doxtator@xxxxxxxxxxxxx wrote:

> message: 2
> date: Tue, 24 Jan 2006 12:54:34 -0500
> from: Don <dr2@xxxxxxxx>
> subject: Re: ILE V OPM Performance Increase?
>
> Doc,
>
> this should be fairly easy to roll-your-own on.  and you could modify it
> to 
> use customer data/apps which makes selling the case alot easier.
>
> Don in DC
>
>
> At 11:41 AM 1/24/2006 -0600, you wrote:
>  
>
> > All
> >
> > Just curious-- has anyone seen any performance improvement in either
> > batch or interactive after converting a program from OPM RPG to ILE
> >    
> RPG?
>  
>
> > (I'm talking about your run-of-the mill batch or interactive programs
> > that do native file I/O, basic calculating and reporting, etc.)
> >
> > If so, are there any numbers that quantify any findings, for either
> >    
> V5R1
>  
>
> > or V5R3?
> >
> > I remember reading several years ago that IBM swore off doing any
> > performance enhancements to the OPM runtime-- I'm wondering if the last
> > few releases have helped the ILE runtime performance.
> >
> > Thanks
> >
> > -Doc
> >    
>
> OK, so there's nothing out there?  Nobody's looked into this?  Nobody
> has even mentioned any anecdotal evidence.  
>
> The reason I ask is this-- I'm trying very hard to push my client into
> ILE.  (Their entire codebase is OPM.) They're fighting with the Cone of
> Fear argument ("We've never done that before."), so I need to have some
> sort of quantifiable argument.  That sort of thing takes time.  I was
> hoping that there was some sort of rudimentary analysis that someone had
> done, just to see if I should even bother with it-- if the speed
> increase is dinky, then there's no point in spending time on it.
>
> Thanks
>
> -Doc
>
>
> This message is for the designated recipient only and may contain privileged, 
> proprietary, or otherwise private information.  If you have received it in 
> error, please notify the sender immediately and delete the original.  Any other 
> use of the email by you is prohibited.
>
>