On Mon, 18 Mar 2013 09:05:57 -0700, Donald J. wrote:
>> I suspect that in your case, the data set was not allocated
>> statically in JCL, as in Monika's case, but dynamically
>> by the temsnote script.
>No, one static dataset. Tivoli situation alerts kick off started
>tasks (all single threaded through same started task name)
>and all use the same .TEMSNOTE z/OS file. $mfile is a 1 line
>single record OMVS file.
>
I still doubt that you test closely resembles Monika's? I have no
idea beyond what you state what Tivoli does. It's easy to
suspect that there more players in the game than you've
describe -- concurrent started tasks, etc.
Are you willing to exhibit the JCL that incurred the error?
Did your JCL contain a static allocation (DD statement) of the
problem data set in a step _after the BPXBATCH step?
Instead, I have simplified Monika's sample JCL:
//
//MONIKA JOB 505303JOB,'Paul Gilmartin',
// MSGLEVEL=(1,1),REGION=0M
//*
//USERC OUTPUT JESDS=ALL,DEFAULT=YES,
// CLASS=R,PAGEDEF=V0648Z,CHARS=GT12
//*
//STEP01 EXEC PGM=BPXBATCH
//STDOUT DD SYSOUT=*
//STDERR DD SYSOUT=*
//STDPARM DD *
SH cp /etc/services "//TEMP.TEST.DATASET"; sleep 2
/*
//STDENV DD *
_BPX_SHAREAS=YES
_BPX_SPAWN_SCRIPT=YES
/*
//STEP02 EXEC PGM=IEFBR14
//SYSPRINT DD SYSOUT=*
//DD01 DD DSN=&SYSUID..TEMP.TEST.DATASET,DISP=SHR
//
:w ! submit $MVS_HOST
The output:
-sh:0+ cp /etc/services //TEMP.TEST.DATASET
cp: FSUM6259 target file "//TEMP.TEST.DATASET": EDC5061I An error occurred when
-sh:0+ sleep 2
I replaced the script with inline commands. Do you believe
that makes a difference?
I used an existing UNIX file instead of creating one. Do you believe
that makes a difference?
If you care to modify the JCL above until it displays the
behavior you report, I'll test it.
Your move,
gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
