Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Mailing Lists -> Oracle-L -> RE: Application architecture & prod/pre-prod switcheroos
"Could you just leave it at changing the synonyms and then remember which
one is pre-prod? You could have a 3rd schema that simply stores the fact
about which of the other schema's is pre-prod and which is prod, and then
use this during the pre-prod loading process."
Yup, that's what we do right now. The loading process has enough checks in it to help me sleep better at night. This is more of a due diligence thing... Is there a more elegant approach with less risk from the DBA side?
"Switching the schemas by changing the synonyms is ingenous. To symplify it, why not have two databases with the same schema - identical except for data content - and just repoint the listener?"
I remember considering that, it seemed to me (at the time) that a synonym drop/recreate could be done a lot more quickly, and with less risk to an established JDBC connection pool, than repointing the listener.
"Or maybe you can have a luxury of having one logon trigger that executes alter session set current_schema?"
Now *that's* ingenious... But again, there's that established connection pool that the J2EE app is using. The sessions are persistent, so login triggers wouldn't work...
Someone else mentioned (privately) using BCV or snap/clone capabilities in the underlying storage in place of the whole Data Pump thing. Not sure what the interplay would be, considering that we are using ASM, but it's another option I'm thinking about.
Appreciate the ideas....
-----Original Message-----
From: Edgar Chupit [mailto:chupit_at_gmail.com]
Sent: Wednesday, January 25, 2006 9:20 AM
To: oracle-l_at_freelists.org
Subject: Re: Application architecture & prod/pre-prod switcheroos
Or maybe you can have a luxury of having one logon trigger that executes alter session set current_schema?
On 1/25/06, Thomas Day <tomday2_at_gmail.com> wrote:
> Switching the schemas by changing the synonyms is ingenous. To
> symplify it, why not have two databases with the same schema -
> identical except for data content - and just repoint the listener?
>
-- Best regards, Edgar Chupit callto://edgar.chupit -- http://www.freelists.org/webpage/oracle-l -- http://www.freelists.org/webpage/oracle-lReceived on Wed Jan 25 2006 - 09:59:38 CST
![]() |
![]() |