Re: Reverse Engineering and selecting specific tables

Hi Hans,
yes we do have lots and lots of detailed technical documentation with
'keep under lock and key' written all over it which describes the base
tables as well as a lot of other documentation. So I guess we have most
of it...

In terms of getting data out of Oracle Apps to include into an EDW we
could either extract from the 'views' and synonyms under the APPS owner
or extract from the underlying base tables.

What we found in a number of cases is the same underlying table
presents itself through a number of different views. HZ_PARTIES being a
good example.....so if we extract from the synonyms/views we get the
same row multiple times and then we actually write code to update the
target table with the relevant fields coming through the different
views/synonyms. If we take the underlying table, acknowledging that it
will be more prone to change in the next release, we can move that data
forward and then actually present the data out of the warehouse using
some of the same view constraints...remembering that we also have
'parties' coming from other systems that will need to be integrated
into the parties coming from oracle...

I see no hard and fast answer.....but I do see a benefit going to the
underlying table...

That said, I must get the client to check with Oracle if there is a
contractual issue with this....I have not heard anyone say that with
regard to such an approach..

I'd be interested to hear what others have done in a similar situation
too.. :-)

Thanks for the feedback...

Peter

.