Re: Job schedulers and activation groups

[Peter Dow <petercdow@xxxxxxxxx>]

Hi Mark,

Thanks for the link. I read it, and the only thing it has to say about 
OVRSCOPE is pretty much the same help text I quoted.  And none of that 
explains the discrepancy between jobs submitted from the Advanced Job 
Scheduler (WRKJOBJS) or the ordinary job scheduler (WRKJOBSCDE).

I guess my question wasn't too clear.  Let me try again.

1) When I submit a job via the old job scheduler, in this case, a 
command CmdA with cpp PgmA which is compiled with *DFTACTGRP, it 
apparently ran in the default activation group.

2) When I submit a job via the advanced job scheduler, the same CmdA 
with cpp PgmA compiled with *DFTACTGRP, it does NOT appear to run in the 
default activation group.

Why?

Scenario 1 seems to be straight forward; the joblog shows a SBMJOB with 
CMD(CmdA).

Scenario 2 is a little more complicated; the joblog shows a SBMJOB with 
CMD(CALL PGM(QIJS/QIJSCRUN) PARM('jobname' '          '  '00' 'QUSRIJS   
' '*LCL    ' '        ' '*LOC      ' '*LOC    '). QIJSCRUN is compiled 
with *DFTACTGRP. It uses "jobname" to get the list of commands for the 
scheduled job, and looks like it uses procedure QIJSC1ALG to execute the 
individual commands, in this case CmdA.  Since the initial program, 
QIJSCRUN, is compiled with *DFTACTGRP, I would expect PgmA to be running 
in the default activation group, but it does not.

*Peter Dow* /
Dow Software Services, Inc.
909 793-9050
pdow@xxxxxxxxxxxxxxx <mailto:pdow@xxxxxxxxxxxxxxx> /


On 3/14/2010 12:26 PM, Mark S. Waterbury wrote:
> Hi, Peter:
>
> Read "Data Management Scoping Rules" (on pp. 48-50 in V5R4, pp. 41-44 in
> V6R1) of the *ILE Concepts manual* in the InfoCenter:
>
>      http://publib.boulder.ibm.com/iseries/
>
> Select V5R4 or V6R1 ... then click on the link for "i5/OS PDF files and
> manuals" in the right pane, and look for the "ILE Concepts" manual.
>
>      V6R1 ILE Concepts
> http://publib.boulder.ibm.com/infocenter/iseries/v6r1m0/topic/books/sc415606.pdf
>
>      V5R4 ILE Concepts
> http://publib.boulder.ibm.com/infocenter/iseries/v5r4/topic/books/sc415606.pdf
>
> HTH,
>
> Mark S. Waterbury
>
>   >  Peter Dow wrote:
>    
> > Hi Rob,
> >
> > I'm no expert on this, but the help for the OVRSCOPE parm says for
> > *ACTGRPDFN "When the activation group is the default activation group,
> > the scope equals the call level of the calling program." and for
> > *CALLLVL "The scope of the override is determined by the current call
> > level.  All open operations done at a call level that is the same as or
> > higher than the current call level are influenced by this override."
> >
> > Which I take to mean that if the program issuing the overrides is in the
> > default activation group (which PgmA is), then it is as if OVRSCOPE was
> > *CALLLVL, and when that's the case, the fact that PgmB is in a new
> > activation group does not matter.
> >
> > *Peter Dow* /
> > Dow Software Services, Inc.
> > 909 793-9050
> > pdow@xxxxxxxxxxxxxxx<mailto:pdow@xxxxxxxxxxxxxxx>  /
> >
> >
> >
> > On 3/13/2010 2:09 PM, rob@xxxxxxxxx wrote:
> >
> >      
> > > Peter,
> > > If PGMA does an override *ACTGRP and then calls PGMB in ACTGRP(*NEW) then
> > > the fact that the override EVER worked is only through the power of some
> > > deity.  But I'll admit, I've never read the ILE Concepts Guide.
> > >
> > >
> > > Rob Berendt
> > >
> > >
> > >