Performance impact of TEST parameters



Support for OS/VS COBOL, VS COBOL II, COBOL for OS/390 & VM and Enterprise COBOL for z/OS

Performance impact of TEST parameters

Postby bravedreamer » Mon Mar 19, 2018 1:48 pm

Hi there,

actually we have activated the compile parameters TEST(NOHOOK,SEPARATE,EJPD),NOOPT in our test-environment for every program that is touched (COBOL compiler 4.2.0, z/os 2.2).
Before placing the program in production we are compiling the code without the parameters.
Obviously this is suboptimal. For one reason, we are putting code in production that is not tested and for the other reason, we are doing work that is not really necessary (recompiling).

So, the question is how big is the performance impact if move the program WITH the test parameters in production?
Is there any improvement if we use a more updated version of the compiler?
bravedreamer
 
Posts: 17
Joined: Wed Nov 11, 2009 6:44 pm
Has thanked: 0 time
Been thanked: 0 time

Re: Performance impact of TEST parameters

Postby prino » Tue Mar 20, 2018 12:24 am

You do not put NOOPT programs in production, ever.
Robert AH Prins
robert.ah.prins @ the.17+Gb.Google thingy
User avatar
prino
 
Posts: 635
Joined: Wed Mar 11, 2009 12:22 am
Location: Vilnius, Lithuania
Has thanked: 3 times
Been thanked: 28 times

Re: Performance impact of TEST parameters

Postby bravedreamer » Tue Mar 20, 2018 8:50 pm

Okay, thanks for the hint.

Actually we have the compiler version 4.2.0. I've seen that in compiler version 5 IBM has changed some parameter.
I thought that maybe the indication of NOOPT is obsolete.
bravedreamer
 
Posts: 17
Joined: Wed Nov 11, 2009 6:44 pm
Has thanked: 0 time
Been thanked: 0 time

Re: Performance impact of TEST parameters

Postby Robert Sample » Tue Mar 20, 2018 9:58 pm

I thought that maybe the indication of NOOPT is obsolete.
Yes and no. NOOPT is deprecated in version 5 of the Enterprise COBOL compiler. HOWEVER, it is replaced by OPTIMIZE(0) which is what is used if you specify NOOPT in this version of the compiler.
Robert Sample
Global moderator
 
Posts: 3719
Joined: Sat Dec 19, 2009 8:32 pm
Location: Dubuque, Iowa, USA
Has thanked: 1 time
Been thanked: 279 times

Re: Performance impact of TEST parameters

Postby Robert Sample » Tue Mar 20, 2018 11:47 pm

For performance of Enterprise COBOL 4.2, have you reviewed http://www-01.ibm.com/support/docview.wss?uid=swg27018287&aid=1 yet? If not, it is a good document to read. Page 8 has
Performance considerations using TEST:
|On the average, TEST(HOOK) was 20% slower than NOTEST, with a range of equivalent to 265%
|slower when not producing a CEEDUMP.

|On the average, TEST(HOHOOK) was equivalent to NOTEST when not producing a CEEDUMP.

|On the average, TEST(NOHOOK,NOSEPARATE) resulted in a 339% increase in the object module
|size compared to using NOTEST or TEST(NOHOOK,SEPARATE), with a range of 158% larger to
|545% larger.

|One program with a large data division (about 1 million items) using TEST(NOHOOK) took 330 times
|more CPU time to produce a CEEDUMP with COBOL's formatted variables compared to using
|NOTEST to produce a CEEDUMP without COBOL's formatted variables
Robert Sample
Global moderator
 
Posts: 3719
Joined: Sat Dec 19, 2009 8:32 pm
Location: Dubuque, Iowa, USA
Has thanked: 1 time
Been thanked: 279 times


Return to IBM Cobol

 


  • Related topics
    Replies
    Views
    Last post