I will have to put an XDC hook in the initialisation exit and poke around.
It looks like the temporary data sets at least have been allocated. We don't
get this problem with systems where 
TEMPDSN is not active in RACF.

An RCF will perhaps clarify the situation but I don't think it will solve my
problem!

Thanks
Robin

-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On
Behalf Of Peter Hunkeler
Sent: 17 September 2018 15:56
To: [email protected]
Subject: Re: ASCRE and ACEE inheritance

 
>Apologies for the delay in replying but your response got flagged as spam
somewhere and I did not see it. I finally found it in the archives! 
 

My bad. The flag had been set by GMX, and I badly missed to remove it from
the subject when replying. I recognized after sending, and for a quick
moment thought to correct and resend, but then decided otherwise. Agrrrr...
I'm sorry.



>I am not sure where the original ACEE of the child STC is coming from since
both it and the parent STC have similar JCL procedure names that match to
the same RACF rule. So the ACEE could be coming from the RACF rule or it
could be inherited (like with forked address spaces or jobs submitted via
internal reader). My problem is that there seem to be two sets of rules in
play: the data sets allocated by JCL get the original STC's ACEE and the
dynamically allocated ones get the new ACEE I set in the address space
initialisation exit (in ASXBSENV). The doc states that the initialisation
exit runs before the JCL procedure but perhaps some JCL processing happens
at the same time. 




I'm surely not authoritative to write "this is how it works", but the
statement "It executes in the new address space before the procedure
identified by SYS1.PROCLIB." Is foggy. A procedure is not executed, programs
are. Your problem makes clear that this part of the documentation should be
enhanced:


- Is the INIT routine run under the region control task, or under the stated
task control task (I suspect the latter)?
- Have any allocations asked for via DD statement in the procedure already
been done? Your observation seems to indicate: yes, they have.


Would an RCF be sufficient?


--
Peter Hunkeler






-- 
Peter Hunkeler 

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to