On Dec 17, 11:12 am, "steven_nospam at Yahoo! Canada"
<steven_nos..._at_yahoo.ca> wrote:
<snip>
>
> Thanks for the advice too. As I said, this was more for me to try and
> understand the logic, but also because this database very rarely
> changes. We did not wnt to implement a standby database setup for it
> (too much overhead), and we could have just exported the data, copied
> it across to the other server, then imported. The thought was that we
> could potentially do a SELECT statement for any entries that are in
> PROD.TABLE1 which are not found in TEST.TABLE1 and do an insert
> statement. With only 8-10 tables in the database, and perhaps 10-20
> new entries per month in "some" of the tables, it could be automated
> to check periodically, and test server would just update itself
> whenever it sees new entries in prod server.
>
> Thx,
>
> Steve
Ah, now that we see the *real* issue ;-), might I suggest you look
into replication. Sounds like that was what you were trying to reinvent.
Very mature technology, has been in Oracle since at least 8.1
that I know of, possibly longer. You can specify source and target
tables, and frequency of replication. Since you are just trying to
keep some small low-activity tables in a test db in synch with a
production, you could set the replication frequency out to once every
24 hours.
Received on Mon Dec 17 2007 - 13:22:11 CST
