Re: Design for backup in mind Was: Backup strategy

[CRPence <crp@xxxxxxxxxxxxxxxxxxxx>]

  My point must have been missed... Restore *does* keep the value from 
the Saved dataspace object; see the suggested modification to the given 
scenario and then review/verify the results.  What was discussed was 
*recovery* activity which is _beyond_ just restore.  There are already 
actions being taken in the given scenario, beyond restore, to effect 
recovery -- the APYJRNCHG itself.  Recovery should be rare.
  I am not saying that storing the updated identity attribute along 
with the insert data can not be done, nor that there would be no value 
in so doing.  In fact I stated that with regard to the NEXT VALUE, it is 
"a valid point that it would be nice if it could be applied" during the 
APYJRNCHG.  However I added that.... In wishing to be able to effect 
that, by applying journaled changes as compared to the relatively 
low-cost method of adding an ALTER RESTART WITH as part of DR, "the 
*cost* is an issue that must be considered".  Cost, as compared to the 
expectation that the value is to be tracked with every [insert] journal 
entry.
  Additionally... Submitting a DCR for inferred as missing function 
*should* be done.  I intended to give extra information about what I 
thought was a resolution to the concern; i.e. a stored procedure to 
effect the change/ALTER after the APYJRNCHG [just like the HA sfw does].

Regards, Chuck
-- All comments provided "as is" with no warranties of any kind 
whatsoever and may not represent positions, strategies, nor views of my 
employer

Walden H. Leverich wrote:
> >   FWiW: The journaled changes are the data.  The attribute of identity
> > value at each insert is not row data, and the value is not separately
> > recorded/tracked neither in its own entry nor as part of the row entry.
>  
> Sorry, I can't really buy that. What you're saying is it's ok that there is no way to "restore" a file to a point in time using the native save/restore methods in i5/OS. Issuing an alter-table isn't a valid part of a save/restore process, IMHO. This is clearly a DCR candidate.
>  
> -Walden