How to force Oracle Streams restart and uni-directional propagation?

From: Joel Lieberman <jlieberman_at_combinatorx.com>
Date: Tue, 5 Feb 2008 09:35:16 -0500
Message-ID: <4A304706170CB7498AC4C49DC3EBACAE0321B412@ares.crx.combinatorx.com>


Greetings -  

We have 2 Oracle database servers - a primary (provider) and secondary (reporting) server. The provider (primary) instance has several schemas that are each duplicated on the secondary (reporting) machine. The replication is one-way only: from the primary to the secondary. After rebooting the machines, the streams infrastructure stopped working (it has restarted without any problems after previous reboots). Capture was in an aborted state for each schema on the provider, and could not be restarted with the OEM GUI interface.  

The alert log shows that the streams cannot find some required archived log files. The application vendor has suggested that we essentially rebuild the entire set of instances and re-setup all of the streams. This seems to be an unnecessarily complex solution.  

From a "business perspective" it appears that we only need to effectively truncate the secondary (reporting) schema, reset the provider streams to capture and propagate and let the one-way flow fill up the secondary schema. I am aware (but not experienced) with the idea that the streams admin schema (STRMADMIN) keeps records of SCNs for all original and replicated objects and uses these in its propagation and application logic. Since we have lost the required logs (I do not know how that could have happened) I am thinking that we should be able to just issue the appropriate commands to the provider and STRMADMIN to tell them to disregard the past and just start up with the current live set of provider and destination schema objects.  

I have not been able to figure out how to force restart for the streams. Should it be as simple as truncating the destination tables and simply forcing a provider restart? What would the DBMS~PACKAGE(s) commands look like? Two days of research have not turned up an actual concrete example of how to get things working again without a complete rebuild.  

Thanks for any help that you can provide -  

Joel E. Lieberman, Ph.D.
Data Systems Engineer
CombinatoRx, Incorporated
245 First Street
Sixteenth Floor
Cambridge, MA 02142  

Voice: 617-301-7058
Email: jlieberman_at_combinatorx.com <mailto:jlieberman_at_combinatorx.com> Company: www.combinatorx.com <http://www.combinatorx.com>

*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*
This email may contain CONFIDENTIAL or PRIVILEGED information and is a private communication between the intended addressee and CombinatoRx, Inc. If you have received this email in error, reading, copying, using, or disclosing its contents to others is prohibited. Please notify us of the delivery error by replying to this message, and then delete it from your system. Thank you.

*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*^*

--
http://www.freelists.org/webpage/oracle-l
Received on Tue Feb 05 2008 - 08:35:16 CST

Original text of this message