Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.server -> Re: Data Warehouse - extract methods
On Sun, 23 Feb 2003 11:28:37 +0000 (UTC), "Dave"
<dave_at_spam-killer-dghconsulting.co.uk> wrote:
>3. Read-only snapshots (materialized views).
>
>Implementing snapshots (materialized views) with fast refreshes would reduce
>the hit on our source systems. However, the initial hit would be large
>(requires an initial full refresh). The snapshot logs would require some
>storage on the source systems, but I wouldn't be too concerned about that if
>we implement regular refreshes. I would foresee us replicating the source
>tables to staging area tables on the data warehouse database. My main
>concern about this option is that it merely moves the change detection
>challenge to the data warehouse database. Change detection is clearly a
>major challenge here.
In the way you want to use mv sure, because you are misusing them. The change detection is *exactly* the thing that is being automated by the snapshot logs, they have the rowids and/or primary keys only. If you are going to 'replicate' (read copy) the source tables to 'stagin' tables, you should seriously ask yourself why you are even willing to do that at all.
Regards
Sybrand Bakker, Senior Oracle DBA
To reply remove -verwijderdit from my e-mail address Received on Sun Feb 23 2003 - 08:35:54 CST
![]() |
![]() |