|
View previous topic :: View next topic
|
| Author |
Message |
abhijit.nayak01
Active User
Joined: 22 Mar 2009
Posts: 161
Location: South Africa
|
|
|
|
Hi,
In my COBOL program the file select is as below:
| Code: |
001800 FILE-CONTROL.
001900 SELECT ATMVALUX ASSIGN TO UT-S-ATMVALUX
001910 FILE STATUS IS W01-ATMBVX-STATUS.
002000 SELECT ATMFICOM ASSIGN TO UT-S-ATMFICOM
002010 FILE STATUS IS W01-ATMFICOM-STATUS. |
There are two scenarios in my JCL and please advise why the JOB is not failing:
Scenario 1:
I am giving wrong file in JCL which doesn't match to COBOL SELECT file i.e. ATMFICOM is not equal to DDDDDDDD in JCL and the job is not abending but its not writing to the output file :
| Code: |
000021 //STP03#06 EXEC PGM=AAAAAAAA
000022 //STEPLIB DD DSN=PPPPPPP.PPPP.LINKLIB,DISP=SHR
000023 //SYSUDUMP DD SYSOUT=*
000024 //SYSABOUT DD SYSOUT=*
000025 //SYSDBOUT DD SYSOUT=*
000026 //SYSPRINT DD SYSOUT=*
000027 //SYSOUT DD SYSOUT=*
000028 //ATMVALUX DD DSN=PPPPPPPP.AAAAAAAA.SRT,DISP=SHR INPUT
000029 //DDDDDDDD DD DSN=BBBBBBB.BBBBBBBB,DISP=OLD OUTPUT |
Scenario 2:
I am not giving any output file JCL which doesn't match to COBOL SELECT file i.e. DDDDDDDD is commented and the job is not abending and it is generating temp file SYS15134.T101419.RA000.MYIDICOM.R0126127 :
| Code: |
000021 //STP03#06 EXEC PGM=AAAAAAAA
000022 //STEPLIB DD DSN=PPPPPPP.PPPP.LINKLIB,DISP=SHR
000023 //SYSUDUMP DD SYSOUT=*
000024 //SYSABOUT DD SYSOUT=*
000025 //SYSDBOUT DD SYSOUT=*
000026 //SYSPRINT DD SYSOUT=*
000027 //SYSOUT DD SYSOUT=*
000028 //ATMVALUX DD DSN=PPPPPPPP.AAAAAAAA.SRT,DISP=SHR INPUT
[color=red]000029 //*DDDDDDDD DD DSN=BBBBBBB.BBBBBBBB,DISP=OLD [/color] OUTPUT |
|
|
| Back to top |
|
 |
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
| It is not clear what you are expecting. You have specified the FILE STATUS clause. If you are not going to test this and cause an abend yourself, just remove it, recompile, rerun, and see if you get what is "expected". |
|
| Back to top |
|
|
Akatsukami
Global Moderator
Joined: 03 Oct 2009
Posts: 1788
Location: Bloomington, IL
|
|
|
|
| Show the code where you are opening ATMFICOM. |
|
| Back to top |
|
|
don.leahy
Active Member
Joined: 06 Jul 2010
Posts: 765
Location: Whitby, ON, Canada
|
|
| Back to top |
|
|
abhijit.nayak01
Active User
Joined: 22 Mar 2009
Posts: 161
Location: South Africa
|
|
|
|
Here is my open file code for output file:
| Code: |
OPEN OUTPUT ATMFICOM.
IF W01-ATMFICOM-STATUS = '00' OR '97'
CONTINUE
ELSE
DISPLAY '*------------------------------------*'
DISPLAY ' INVALID FILE STATUS WHILE OPENING THE '
DISPLAY ' INPUT ATMFICOM FILE IN PGM ATMBICOM '
DISPLAY ' IN SECTION : B-INITIALISE '
DISPLAY ' FILE STATUS : ' W01-ATMFICOM-STATUS
DISPLAY '*------------------------------------*'
PERFORM Z999-ABEND
END-IF.
|
|
|
| Back to top |
|
|
abhijit.nayak01
Active User
Joined: 22 Mar 2009
Posts: 161
Location: South Africa
|
|
|
|
Hi Don,
Where can I find CBLQDA option:
The below is the area where I am compiling via chageman and I am not sure how to make CBLQDA OFF.
| Code: |
PACKAGE ID: ****002515 STATUS: DEV INSTALL DATE: 20150530
STAGED NAME: ATMBICOM
LIBRARY TYPE: SRC - Source Library - Batch
DATASET NAME: *******.***.STG.#002515.SRC
LANGUAGE ===> COB322 (Blank for list; applies to source code)
COMPILE PROCEDURE ===> COB322 (Blank for list; ? for designated procedure)
COMPILE PARMS ===>
LINK EDIT PARMS ===> AMODE(31)
DB2 PRECOMPILE ===> NO (Y/N) PRECOMPILE VARIABLES ===> NO (Y/N)
OTHER OPTIONS ===> YES (Y/N to display other options)
JOB STATEMENT INFORMATION:
===> //****ICOM JOB (999X999,DS),****,CLASS=C,MSGCLASS=V,NOTIFY=****_________
===> //*_____________________________________________________________________
===> //*_____________________________________________________________________
===> /*JOBPARM SYSAFF=*______________________________________________________ |
|
|
| Back to top |
|
|
abhijit.nayak01
Active User
Joined: 22 Mar 2009
Posts: 161
Location: South Africa
|
|
|
|
Hi Bill,
I removed the file status and compiled and reran the job but the job still ran fine. |
|
| Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
Don may have a very good shot.
Include this is your step:
| Code: |
//CEEOPTS DD *
RPTOPTS(ON) |
This will list the values of all the Language Environment options for your run.
Check the value of CBLQDA which is shown to you. If it is (ON), then use this:
| Code: |
//CEEOPTS DD *
CBLQDA(OFF)
RPTOPTS(ON) |
|
|
| Back to top |
|
|
don.leahy
Active Member
Joined: 06 Jul 2010
Posts: 765
Location: Whitby, ON, Canada
|
|
|
|
CBLQDA is an LE option, not a Cobol compiler option. To see the list of options in use in your shop, ask your friendly systems programmer.
Or add the following to your JCL:
| Code: |
//OPTRPRT DD SYSOUT=*
//CEEOPTS DD *
RPTOPTS(ON),MSGFILE(OPTRPRT)
/* |
|
|
| Back to top |
|
|
abhijit.nayak01
Active User
Joined: 22 Mar 2009
Posts: 161
Location: South Africa
|
|
|
|
Hi,
I found CBLQDA is ON . Kindly see below:
LAST WHERE SET OPTION
| Code: |
-------------------------------------------------------------------------------
PARMLIB(CEEPRM02) ABPERC(NONE)
PARMLIB(CEEPRM02) ABTERMENC(ABEND)
PARMLIB(CEEPRM02) NOAIXBLD
PARMLIB(CEEPRM02) ALL31(OFF)
PARMLIB(CEEPRM02) ANYHEAP(16384,8192,ANYWHERE,FREE)
PARMLIB(CEEPRM02) NOAUTOTASK
PARMLIB(CEEPRM02) BELOWHEAP(8192,4096,FREE)
PARMLIB(CEEPRM02) CBLOPTS(ON)
PARMLIB(CEEPRM02) CBLPSHPOP(OFF)
PARMLIB(CEEPRM02) CBLQDA(ON)
PARMLIB(CEEPRM02) CEEDUMP(60,SYSOUT=*,FREE=END,SPIN=UNALLOC)
PARMLIB(CEEPRM02) CHECK(ON)
PARMLIB(CEEPRM02) COUNTRY(ZA) |
and after I gave
| Code: |
//CEEOPTS DD *
CBLQDA(OFF)
RPTOPTS(ON)
/* |
The job abended which is expected but can't it be done something in COBOL program in code or handled inside COBOL code. I mean if I am checking file status then why it didn't picked 96?:
| Code: |
*------------------------------------*
INVALID FILE STATUS WHILE OPENING THE
INPUT ATMFICOM FILE IN PGM ATMBICOM
IN SECTION : B-INITIALISE
FILE STATUS : 96
*------------------------------------* |
|
|
| Back to top |
|
|
don.leahy
Active Member
Joined: 06 Jul 2010
Posts: 765
Location: Whitby, ON, Canada
|
|
|
|
Well, according to the 2002 Cobol Standard, the behavior that you see with CBLQDA(ON) is "correct". According to the Standard, a conforming program should not fail due in this situation, it should dynamically create a file.
CBLQDA(OFF) is the IBM supplied default, so someone at your shop must have chosen to override it. I suggest that you ask the systems programming team that is responsible for LE why they made that decision. |
|
| Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
Yes, it's even earlier than the 2002 (now 2014) Standard, it goes back to the 85 Standard.
This is from the documentation for CBLQDA:
| Code: |
ON
Specifies that COBOL QSAM dynamic allocation is permitted. ON conforms to the 1985 COBOL standard. |
If your technical staff change the default, you may get countless program failures. Might be a good idea, because those programs can hardly be "working" if they are just losing particular outputs :-) |
|
| Back to top |
|
|
don.leahy
Active Member
Joined: 06 Jul 2010
Posts: 765
Location: Whitby, ON, Canada
|
|
|
|
| I first learned about this issue when I helped out a programmer who'd spent 2 days banging his head against the wall trying to figure out why his output file wasn't being populated. He had ruled out a misspelled ddname because he expected that it would have caused an obvious error. And it would have, a few weeks before, when the setting had been CBLQDA(OFF). Someone on the systems programming staff had recently switched it to ON during a maintenance activity. |
|
| Back to top |
|
|
abhijit.nayak01
Active User
Joined: 22 Mar 2009
Posts: 161
Location: South Africa
|
|
|
|
Well this is something a dangerous thing and also I suffered a serious problem ten days before when I changed an existing program and the requirement was to add one more output file to the program. I did that part and in JCL I added the new output file but somehow I tampered the the existing output file in JCL. The job was implemented in production and for me the job ran fine as I checked my new output file and I checked the existing output file which had already previous data . After three days customer complained they were receiving same data from past three days and all the settlement are messed up. Then I found that there was a mistake in JCL but I couldn't understand why this happened as the job should have abended as per my basic JCL knowledge. Then it took 4 days to recover the data and get the things settled. One my colleague said may be I didn't check for file status in the COBOL program that's why it didn't fail. Then today I created a small program and gave the file status to check whether it abends or not and the result is in the previous posts. Will check with system guys for the default CBLQDA.
Thanks guys for clarifying the doubt. |
|
| Back to top |
|
|
Robert Sample
Global Moderator
Joined: 06 Jun 2008
Posts: 8696
Location: Dubuque, Iowa, USA
|
|
|
|
| Quote: |
| as the job should have abended as per my basic JCL knowledge. |
I think this is the real issue. Your problem is not JCL, it is COBOL. For a utility, a missing DD statement will cause an ABEND or non-zero return code. However, as you have discovered, COBOL will -- in some cases -- run a program even if a file DD name is not found. The Enterprise COBOL Programming Guide makes clear that in such a case the allocated file persists only for the length of the run unit -- that is, when your batch job completes the file goes away (it is a temporary file only).
There are a few cases where I've used BPXWDYN and executed COBOL programs in batch with NO DD statements. This is not a common need, but the code can be written. |
|
| Back to top |
|
|
|
|
