Comments on: Mview Complete Refresh and Atomic_refresh parameter http://askdba.org/weblog/2009/07/mview-complete-refresh-and-atomic_refresh-parameter/ Writing About Our Experiences With Oracle Databases Fri, 05 Mar 2010 10:24:30 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Amit http://askdba.org/weblog/2009/07/mview-complete-refresh-and-atomic_refresh-parameter/comment-page-1/#comment-4365 Amit Fri, 24 Jul 2009 05:37:55 +0000 http://askdba.org/weblog/?p=500#comment-4365 Chris, Thanks for your comments. Yes, this issue is because of bug 3168840. But I faced the issue as a result of bug fix :), where in atomic refresh for a large mview (~195 Gb) has generated lot of undo and had not completed deletion of data after more then 12 hours. I believe atomic_refresh=true parameter should be used judiciously and avoided for large mviews. Regards Amit Chris,

Thanks for your comments. Yes, this issue is because of bug 3168840. But I faced the issue as a result of bug fix :) , where in atomic refresh for a large mview (~195 Gb) has generated lot of undo and had not completed deletion of data after more then 12 hours. I believe atomic_refresh=true parameter should be used judiciously and avoided for large mviews.

Regards
Amit

]]> By: Christian Antognini http://askdba.org/weblog/2009/07/mview-complete-refresh-and-atomic_refresh-parameter/comment-page-1/#comment-4362 Christian Antognini Thu, 23 Jul 2009 21:26:55 +0000 http://askdba.org/weblog/?p=500#comment-4362 Hi The problem related to atomic_refresh is a bug. At page 474 in TOP (http://antognini.ch/top) I wrote the following text about this topic: "In Oracle9i, because of bug 3168840, even if the parameter atomic_refresh is set to TRUE (the default in all versions), during a complete refresh of a single materialized view, a TRUNCATE statement is executed. If several materialized views are refreshed at the same time however, the refresh works atomically as expected. Therefore, to work around this bug, it is possible to create a refresh group containing the materialized view you have to refresh and a “dummy” materialized view that is created only to have a second one." Cheers, Chris Hi

The problem related to atomic_refresh is a bug. At page 474 in TOP (http://antognini.ch/top) I wrote the following text about this topic:

“In Oracle9i, because of bug 3168840, even if the parameter atomic_refresh is set to TRUE (the default in all versions), during a complete refresh of a single materialized view, a TRUNCATE statement is executed. If several materialized views are refreshed at the same time however, the refresh works atomically as expected. Therefore, to work around this bug, it is possible to create a refresh group containing the materialized view you have to refresh and a “dummy” materialized view that is created only to have a second one.”

Cheers,
Chris

]]> By: Blogroll Report 03/07/2009 – 10/07/2006 « Coskan’s Approach to Oracle http://askdba.org/weblog/2009/07/mview-complete-refresh-and-atomic_refresh-parameter/comment-page-1/#comment-3849 Blogroll Report 03/07/2009 – 10/07/2006 « Coskan’s Approach to Oracle Fri, 10 Jul 2009 18:42:06 +0000 http://askdba.org/weblog/?p=500#comment-3849 [...] Amit Bansal – Mview Complete Refresh and Atomic_refresh parameter [...] [...] Amit Bansal – Mview Complete Refresh and Atomic_refresh parameter [...]

]]>