Comments on: Make History with Streams http://www.oraclealchemist.com/oracle/make-history-with-streams/ Steve Karam's Oracle Blog Thu, 29 Jul 2010 20:12:51 +0000 http://wordpress.org/?v=2.9.1 hourly 1 By: My 1st Oracle Stream experience . | http://www.oraclealchemist.com/oracle/make-history-with-streams/comment-page-1/#comment-44320 My 1st Oracle Stream experience . | Mon, 23 Nov 2009 17:29:39 +0000 http://www.oraclealchemist.com/oracle/make-history-with-streams/#comment-44320 [...] Custom Insert Handler for Oracle Stream  on Steve Karam [...] [...] Custom Insert Handler for Oracle Stream  on Steve Karam [...]

]]> By: BRR http://www.oraclealchemist.com/oracle/make-history-with-streams/comment-page-1/#comment-44296 BRR Wed, 02 Sep 2009 18:57:47 +0000 http://www.oraclealchemist.com/oracle/make-history-with-streams/#comment-44296 The script works for tables which have single column primary keys. If there are more then one column in primary key how can we handle? and if primary key is of char type? Thanks The script works for tables which have single column primary keys.
If there are more then one column in primary key how can we handle?
and if primary key is of char type?

Thanks

]]> By: JAYT http://www.oraclealchemist.com/oracle/make-history-with-streams/comment-page-1/#comment-44073 JAYT Wed, 06 May 2009 12:23:48 +0000 http://www.oraclealchemist.com/oracle/make-history-with-streams/#comment-44073 I set a similar setup up for a BI application collecting data from nationwide systems last year. The volumes were high though. Although the DML handler was effficient in itself, it added an overhead of 100ms to each row. This was too much for our data volumes. Eventually dropped the DML handler and scheduled a delete on the source tables to delete any data more than a specified number of days old. In fact, as Streams picks up data from redo logs, could simply delete data as soon as it is inserted, as long as deletes are not propagated. We retained data for three days for verification purposes only. I set a similar setup up for a BI application collecting data from nationwide systems last year.

The volumes were high though.

Although the DML handler was effficient in itself, it added an overhead of 100ms to each row. This was too much for our data volumes.

Eventually dropped the DML handler and scheduled a delete on the source tables to delete any data more than a specified number of days old. In fact, as Streams picks up data from redo logs, could simply delete data as soon as it is inserted, as long as deletes are not propagated. We retained data for three days for verification purposes only.

]]> By: Eric Choong http://www.oraclealchemist.com/oracle/make-history-with-streams/comment-page-1/#comment-44012 Eric Choong Thu, 20 Nov 2008 09:55:51 +0000 http://www.oraclealchemist.com/oracle/make-history-with-streams/#comment-44012 Hi Steve, I have setup a streams environment similar to above. My source and destination schemas are on the same database hence i do not have a propagate rules. Both capture and apply rules are using the same queue. My problem is, I'm not able to replicate DDL. DML is fine. Do you have any idea? Thanks and regards, Eric C Hi Steve,
I have setup a streams environment similar to above. My source and destination schemas are on the same database hence i do not have a propagate rules. Both capture and apply rules are using the same queue. My problem is, I’m not able to replicate DDL. DML is fine. Do you have any idea?

Thanks and regards,
Eric C

]]> By: Govind Thanavelu http://www.oraclealchemist.com/oracle/make-history-with-streams/comment-page-1/#comment-43975 Govind Thanavelu Wed, 01 Oct 2008 18:55:44 +0000 http://www.oraclealchemist.com/oracle/make-history-with-streams/#comment-43975 It is useful info. Thank you. It is useful info. Thank you.

]]> By: Gyongyi http://www.oraclealchemist.com/oracle/make-history-with-streams/comment-page-1/#comment-43921 Gyongyi Mon, 14 Jul 2008 20:54:02 +0000 http://www.oraclealchemist.com/oracle/make-history-with-streams/#comment-43921 Hi, I would like to ask you about the possibility apply the same handler for several tables (whole schema ) without specifying the DBMS_APPLY_ADM.SET_DML_HANDLER for each of the table. My situation is: I would like to replicate all tables with additional 3 columns (timestamp, user, dml operation). This transformation is done in one plsql proc - the handler. However in the apply phase I have to define for each db object (tables) three set_dml_handler (update, insert, delete) statment point to the handler. And this is exactly I would like avoid, the workload to define 3 set_dml_handlers per tables (in my shcema is around 1000 tables). Can you advice how to manage it in easier way ? We have 10g DB. Many thanks, GG Hi,
I would like to ask you about the possibility apply the same handler for several tables (whole schema ) without specifying the DBMS_APPLY_ADM.SET_DML_HANDLER for each of the table.
My situation is: I would like to replicate all tables with additional 3 columns (timestamp, user, dml operation). This transformation is done in one plsql proc – the handler. However in the apply phase I have to define for each db object (tables) three set_dml_handler (update, insert, delete) statment point to the handler. And this is exactly I would like avoid, the workload to define 3 set_dml_handlers per tables (in my shcema is around 1000 tables). Can you advice how to manage it in easier way ?
We have 10g DB.
Many thanks,
GG

]]> By: sujit http://www.oraclealchemist.com/oracle/make-history-with-streams/comment-page-1/#comment-43433 sujit Wed, 04 Jun 2008 15:03:22 +0000 http://www.oraclealchemist.com/oracle/make-history-with-streams/#comment-43433 Hello there, Can you please send me steps required to replicate same setup which you have done? We need something similar to this. Regards, Sujit Hello there,

Can you please send me steps required to replicate same setup which you have done?

We need something similar to this.

Regards,
Sujit

]]>