From the JCL:
//P PROC S='no sub'
//PSTEP EXEC PGM=BPXBATCH,PARMDD=PARMDD
//PARMDD DD *,SYMBOLS=CNVTSYS
SH set -x; sleep 11;
echo "Hello, world! from PARMDD. &&S &S";
echo "Time is &HR:&MIN:&SEC Date is &YR4-&MON-&DAY.."
//STDOUT DD SYSOUT=(,)
//STDERR DD SYSOUT=(,)
//STDIN DD PATH='/dev/null'
// PEND
//*
//D1 EXEC PGM=IEFBR14 (Timestamp in job log.)
//STEP1 EXEC P,S='Substituted in step 1'
//D2 EXEC PGM=IEFBR14 (Timestamp in job log.)
//STEP2 EXEC P,S='Substituted in step 2'
... the job long shows:
08.17.51 JOB04792 - -----TIMINGS
(MINS.)------
08.17.51 JOB04792 -STEPNAME PROCSTEP RC EXCP CONN TCB SRB
CLOCK
08.17.51 JOB04792 -D1 00 8 2 .00 .00
.0
08.18.02 JOB04792 -STEP1 PSTEP 00 60 9 .00 .00
.1
08.18.02 JOB04792 -D2 00 10 2 .00 .00
.0
08.18.13 JOB04792 -STEP2 PSTEP 00 62 9 .00 .00
.1
08.18.13 JOB04792 IEF404I PARMDD - ENDED - TIME=08.18.13
... and in the respective STDOUTs:
********************************* TOP OF DATA *****************
Hello, world! from PARMDD. &S Substituted in step 1
Time is 14:17:51 Date is 2014-08-15.
********************************* TOP OF DATA *****************
Hello, world! from PARMDD. &S Substituted in step 2
Time is 14:18:02 Date is 2014-08-15.
******************************** BOTTOM OF DATA ***************
The substituted times match the step start times. I coded
"SYMBOLS=CNVTSYS". Does the converting system somehow know
when the steps will run on the executing system? I doubt it.
Does the executing system somehow reach back to the converting
system to extract the instantaneous values of its system
symbols? In this case, the executing system is the same as
the converting system. Would the results be different if two
different systems were involved?
What happens if the job runs very close to midnight? Might
the time be extracted in the previous day and the day in the
following, resulting in a 23:59:59 error? Or is an atomic
snapshot of dynamic system symbols taken and used for the
duration of the step?
I understand the reason for the extra spaces between the Time
and the Date; WAD. I still don't like the behavior.
What are the multiple:
10 ++S EXPORT EXPSET=Substituted in step 1 GENERATED
STATEMENT
... I see unexpectedly in my JESJCL? (No message code!?)
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
