Re: QAQQINI Veracity

[CRPence <CRPbottle@xxxxxxxxx>]

  There would normally be only the one copy of the file QAQQINI in 
QSYS.  That is the /model/ copy of the file, with default data.  Other 
copies of that file would exist when desired initialization options, 
other than the defaults, should be activated.  A copy of QAQQINI in 
QUSRSYS serves as an effective global default, except where CHGQRYA has 
been used to specify a different QRYOPTLIB() [query options library] 
/other than/ QUSRSYS.

  If all of the values are the default, then there is reason to *not* 
have a copy in QUSRSYS.  Many times the file is created only to get rid 
of the message that the file was not found; seen with service\debug 
active.  That is somewhat pointless, as then additional work dealing 
with the options [file] is performed, and the message that was since 
eliminated is replaced by another message which states from where the 
options were loaded.

  Only each *copy* of the file that will be activated for 
initialization options should be changed to specify the desired modified 
options.  The copy of QAQQINI in QSYS should fail for attempts to 
changes its data, as enforced by the triggers.  The triggers must also 
be copied when creating a copy of the model:
   CRTDUPOBJ QAQQINI QSYS *FILE MyLib *SAME DATA(*YES) TRG(*YES)

Regards, Chuck

Jerry Adams wrote:
> I installed the latest V5R4 Cume yesterday, and one of the special 
> instruction cover letters noted that to take advantage of the PTF it 
> would be a good idea to update the Parallel_Degree parameter.  All fine 
> and good, but I noticed that there are two [2] QAQQINI files on our 
> System i5.
>
> So I went to the Midrange archives for help and solace.  So I found that 
> there usually are two versions of the table; one in QSYS and the "real" 
> one in QUSRSYS.  I gathered that the one in QUSRSYS is the one used by 
> most (all?) system functions.  But I noticed that the one in QSYS had 
> more parameters defined (30) than the one in QUSRSYS (23).  I suspect 
> that the reason for the difference is the V5R3 -to- V5R4 upgrade we did 
> last year.  But the new version (QSYS), I'm surmising, didn't get 
> replicated to QUSRSYS.
>
> OK so far.  My question is: Can I simply delete the version in QUSRSYS 
> and replicate (using CRTDUPOBJ) the one in QSYS to QUSRSYS?  In practice 
> it probably doesn't matter since all of the parameters in both versions 
> have a value of *Default, but I'm a neat freak (don't look at my desk, 
> though).
>
> Also, does it even matter if the parameter values are ever updated in 
> QSYS?  Or do I need to update both versions whenever I change a 
> parameter's value?