IBM Mainframe Forum

Hi Friends,

I am compiling cobol elements using ENDEVOR and trying to get the compilation report in a PS file (Txx.Txxxxx.ENDV.RPT1)by updating JCL as below:



Somehow, getting below JCL error



Can you please help to get what could be the probable issue with this JCL

The Messages and Codes manual is fairly explicit on this. What did you not understand?

This is not a JCL error! The IEC141I message is issued by the OPEN routine.

RC 20 means OPEN recognizes BLKSIZE is not an integral multiple of LRECL. You specify LRECL=134,BLKSIZE=2680, and here BLKSIZE is 20 * LRECL. So OPEN seems to be wrong at first, but it surely is not. What's going on here, you might ask?

DCB attributes used by OPEN are merged from 3 sources: 1. DCB in program, 2. JCL, 3. Attibute of the data set (if it exists alredy). Whatever is specified in the DCB by the programmer (ENDEVOR in this case) cannot be overwritten by sources 2.) or 3.) above. The programmer usually writes its code to work with either fixed or variable length records of a specific length. So RECFM, and LREC are commonly defined by the program code. You cannot overwrite the by sources 2., or 3. For teh block size (BLKSIZE) hoever, it is good programming practise not to specify is in the code but leave it undetermined until runtime. So, you usually can set it via sources 2., or 3.

The net of all this: I bet ENDEVOR specifies that the C1MSG2 data set must have attributes RECFM=F [ B ] A, and either LRECL=133, or LRECL=121. You are setting LRECL=134, which is being ignored by OPEN, and BLKSIZE 2680, which is being accepted by OPEN. But 2680 is not an integral multiple of neither 133, nor 121.

Try with removing all of REFCM=, LRECL=, and BLKSIZE= from your DD statement (and delete the data set if it exists, or use a new one), and see what happends. Alternatively, if you insist in specifyin BLKSIZE, use an integeral multiple of whatever LRECL the ENDEVOR manual tells you (as said, I would assume 133, or 121).

Thanks phunsoft,

After updating LRECL = 133, JCL worked fine for me. I have one more query here. Once this JOB went fine, one more JOB submitted via internal reader for element movement to stage 2. The 2nd JCL also contain C1MSGS2 DD statement, and I want to use PS file instead of SYSOUT=*, just like I used in first JCL.

Basically I need to have all element movement details in a PS file for reporting. Is there any way we can override the C1MSGS2 DD statement in 2nd JCL. ??

Not sure I understand what your problem is. To have the output be written to a data set instead of to the JES spool, you need to modify the JCL before it is submitted, or written to the internal reader. Just as you did with your first job.

Endevor keeps a history of all movements so you do not need to. It will also write its processing output to a data set (VSAM, I think) that cn be queried through the Endevor menu system. If necessary, talk to your Endevor administrator.

As this topic has been completed I will lock it.