Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.misc -> Re: Oracle Transaction Boundary ..
Comments in-line.
Raghavendra S. Kothamangala wrote:
> Hi All,
>
> Thanks for the replies.
>
> Here is more information about the requirement.
>
> I have a lot of J2EE applications modifying the oracle database.
> The data which is updated in the oracle database has to be replicated
> in another database (non-oracle).
But you won't tell us the name of the product or its version number? Nor do you refer to it again in anything else you have written.
> Before doing the update, I have to transform the data, and make sure
> I maintain the transaction integrity.
> Constraints:
> 1. I cannot change the existing J2EE applications.
> 2. Need to rollback change, if one table does not go through.
> Transaction Integrity.
Rollback what change? In which database? For what reason might it not "go through"? What does "go through" mean? Are the transactions inserts? updates? deletes?
> I have ruled out CDC (Change Data Capture in 10G) as we need to use
> Views for achieving the same.
Why do you "need" to use views?
> I am exploring using the dbms_transactions package.
> We might have to have triggers on all tables in the database which can
> get modified by J2EE apps.
Use a J2EE app to modify triggers or to modify the data?
A few lines earlier you said you couldn't change the J2EE app. Here's precisely what you said: "I cannot change the existing J2EE applications." So which is it? You can or you can't?
> When there is modification to the database, we will need to write the
> data into a separate table common for all the tables in the database,
> and then take the data from there.
>
> Thanks again
> -Raghu
I understand that English is most likely not your first or second language but your statements above are either self-conflicting or ambiguous to the point of having no meaning?
If you need help with something you need to very carefully define what it is and how/where someone can offer you help. At this point: No help is possible.
-- Daniel A. Morgan University of Washington damorgan_at_x.washington.edu (replace 'x' with 'u' to respond)Received on Sat May 14 2005 - 18:36:39 CDT