Re: Index Rebuild

[Jim Oberholtzer <midrangel@xxxxxxxxxxxxxxxxx>]

Another thing to check to be sure is that you don't have cross attached 
logical files.  Meaning a logical that is in "TEST" while pointing to 
"PROD" data.  I can't count how many times I have pointed out to folks 
that they have created these crossed logicals in the test environment 
and they steadfastly deny it was done.  The database always says the 
truth when asked about these types of references.

On a restore the logical would rebuild, so while I would follow Luis' 
advice, I would check the cross attachments as well.

Jim Oberholtzer
Chief Technical Architect
Agile Technology Architects


On 4/12/2012 12:45 PM, Luis Rodriguez wrote:
> Jerry,
>
> Check the value of the ACCPTH parameter in your SAVOBJ or SAVLIB command.
>
> Regards,
>
> Luis Rodriguez
> IBM Certified Systems Expert — eServer i5 iSeries
> -- On Thu, Apr 12, 2012 at 1:10 PM, Jerry C. Adams <midrange@xxxxxxxx> 
> wrote:
> > >  First, we're on V5R1.  So this may be (oh, heck, IS) an issue.
> > >
> > >
> > >
> > >  Every night I save a copy of our data library to a test library, which I
> > >  can
> > >  use for testing (no partitions).  The test library is then saved to a save
> > >  file.  I can restore the test library from the save file, if my tests,
> > >  well,
> > >  screw up.
> > >
> > >
> > >
> > >  I don't often have to restore it in the middle of the day, but it does
> > >  happen.  In fact, today I've done it twice (bad day).  Obviously things go
> > >  a
> > >  little slow (to put it mildly) during the restore.  But even after my
> > >  command returns to the menu, there are still jobs running that are building
> > >  indices.  WRKACTJOB shows:
> > >
> > >
> > >
> > >  Subsystem/Job  User        Type  CPU %  Function        Status
> > >
> > >    QDBSRV05     QSYS        SYS    46.8  IDX-A.HISLA1     RUN
> > >
> > >    QDBSRV04     QSYS        SYS    43.3  IDX-D.LYMDRL01   RUN
> > >
> > >
> > >
> > >  The restore is done by, first, clearing the library.  Then doing a RSTOBJ
> > >  to
> > >  the test library.  Is there a better, quicker way to do this?  I never
> > >  noticed this before (i.e., on other systems on which I have worked using
> > >  the
> > >  same technique), but I would have thought (erroneously it seems) that the
> > >  indices would have simply been restored from the save file, instead of
> > >  under-going a complete rebuild.  The LFs are in the save file.  The disks
> > >  are mirrored.
> > >
> > >
> > >
> > >  Jerry C. Adams
> > >
> > >  IBM i Programmer/Analyst
> > >
> > >  You'll never make it - four groups are out.  Go back to Liverpool. - Decca
> > >  Records executive to the Beatles (1962)
> > >
> > >  --
> > >
> > >  A&K Wholesale
> > >
> > >  Murfreesboro, TN
> > >
> > >  615-867-5070
> > >
> > >
> > >
> > >  --