I've been struggling with this one for a couple of days now, without making
any headway.

I have a REXX that works through SDSF that is intended to issue 'D R,L'
and retrieve any outstanding console messages. It works fine when I invoke
it from ISPF option 6 with an EXEC... command, but when I invoke it via a
batch job or a started task it doesn't return the expected results.

After a considerable amount of manual bashing and web browsing, I've found
a couple of threads describing a similar problem on various web fora, the
most notable one having been raised back in 2011, and responded to by a
number of the regular contributors here on IBM-MAIN.

Unfortunately, I can't make the suggested modifications do anything for me,
and I'm hoping that someone here might be able to point me in a different
direction.

My REXX is:
/* REXX */
rc=isfcalls('ON')
rc=syscalls('ON')
isfcons="NODDY"
Address SDSF ISFEXEC "'/D R,L' (wait"
say "isfulog.0: "isfulog.0
if isfulog.0>0 then do
  do i=1 to isfulog.0
    say i isfulog.i
  end
end
rc=isfcalls('OFF')
rc=syscalls('OFF')

When I run this via EXEC... in option 6, the result is:
isfulog.0: 6
1 S0W1      2019150  03:02:03.62             ISF031I CONSOLE NODDY ACTIVATED
2 S0W1      2019150  03:02:03.62            -D R,L
3 S0W1      2019150  03:02:03.62             IEE112I 03.02.03 PENDING
REQUESTS 774
4                                            RM=1    IM=0     CEM=0
EM=0     RU=0    IR=0    NOAMRF
5                                            ID:R/K     T SYSNAME  MESSAGE
TEXT
6                                                    38 R S0W1     *38
DFS996I *IMS READY*  IVP1
… which is exactly as expected.

If I use a console START command to invoke the REXX via some started task
JCL...
//RUNEXEC  EXEC PGM=IKJEFT01,PARM='%<execname>'
//SYSEXEC  DD   DSN=REXXLIB.SYSEXEC,DISP=SHR
//SYSTSPRT DD   SYSOUT=*
//SYSTSIN  DD   DUMMY

...the SYSTSPRT output on the hold queue is;
IKJ56644I NO VALID TSO USERID, DEFAULT USER ATTRIBUTES USED
isfulog.0: 0
READY
END

As can be seen, the isfulog stem has simply gone west.
The same occurs if I modify the JCL to turn it into a batch job and invoke
the REXX that way instead.
I've tried a large number of variants of the REXX logic to trace it's
progress, and in the hope of forcing something different to happen, all to
no avail.

Any ideas, anyone... please?

Regards
Sean

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

Reply via email to