[Profile Register/Log In] (What's This?)

From: Carol Dacko <cdacko@umich.edu>
Subject: [techies] Re: CDC
In-Reply-To: [techies] CDC
Date: Fri Jan 8 17:20:41 2016

CDC is a supported option in Oracle but there is no development on this product anymore. All efforts are being placed into the GoldenGate product. 

With our CDC installation (12 tables in production) we still have instances where the apply process just fails with ORA-600 errors. All efforts to try to get this resolved with MOS was only partial successful and after many patch applications. After months of dealing with MOS, we decided we would deal with the occasional failure and restart the failed process. We also know that CDC will not be a long-term solution for us! 

We use asynchronous hot-log mode of CDC. And for each table where CDC is enabled, there are 7 or 8 processes that get started with it in this mode. I don't know if this is for all modes of CDC.

If it were me and it was needed for many years in the future, I would look at different Change Data Capture product including GoldenGate or whatever other products that are on the market.

BUT, it is free and you can't beat that!


Carol Dacko
University of Michigan

On Wed, Apr 18, 2012 at 8:40 PM, Sorrin, Eric <Eric.Sorrin@childrenscolorado.org> wrote:

(Mailing list information, including how to remove yourself, is located at the end of this message.)



We are looking into using CDC to populate our data warehouse.  We would like to minimize impact to our source database, which currently is not in archivelogmode.


I’ve been staring at Oracle documentation and web posts for days, and find the information confusing and conflicting.


Specifically, one source shows that database links are necessary with all but synchronous and Asynchronous Hotlog – so needed for Asynchronous Distributed Hotlog, Asynchronous Autolog online, and Asynchronous Autolog archive.  However, the Oracle documentation shows DB link only for Asynchronous Distributed Hotlog.


So, Question 1, when are db links necessary?


Next question regards archivelogmode.  We really do not want to put the source in archivelogmode.  It would seem that Asynchronous Autolog Online is the way to go to minimize impact to the source system, but also looks complicated with lots of moving pieces.  Furthermore, the Oracle documentation states “It is assumed that the database is currently running in ARCHIVELOG mode”.


So, question 2, when is it mandatory that the source DB be in archivelogmode?  It’s not needed for synchronous, obviously needed for Asynchronous Autolog archive, not so clear for the rest.


Question 3.    This regards data integrity.  What happens in all of the cases where the staging/destination reside somewhere else (Asynchronous Distributed Hotlog, Asynchronous Autolog Online, Asynchronous Autolog archive) and that database is down?  Are all of these options viable ways to ensure we will not lose data if the source is up and the destination is down?


Hopefully someone out there has struggled through this already.


Thanks in advance,


Eric Sorrin


Children’s Hospital Colorado

CONFIDENTIALITY NOTICE: This e-mail is confidential, may be legally privileged,
and for the intended recipient only. Access, disclosure, copying, forwarding and
distribution by any means is strictly prohibited. If received in error,
do not read but delete and e-mail confirmation to the sender.


Subscription Reminder: You're Subscribed to: techies using the address: cdacko@umich.edu

From: heidikuhn@rmoug.org
P.O. Box 621942
Littleton, CO 80162

Manage Your Subscription » or, Unsubscribe Automatically »

Mailing List Powered by Dada Mail