Comments on: 3 ways to maintain rollup tables in SQL http://www.xaprb.com/blog/2006/07/19/3-ways-to-maintain-rollup-tables-in-sql/ Stay curious! Thu, 02 May 2013 12:36:53 +0000 hourly 1 http://wordpress.org/?v=3.5.1 By: gavis http://www.xaprb.com/blog/2006/07/19/3-ways-to-maintain-rollup-tables-in-sql/#comment-14143 gavis Thu, 10 Jan 2008 07:59:31 +0000 http://www.xaprb.com/blog/?p=161#comment-14143 how do i go about maintaining 300 databases which each have 100 tables, to ensure the best performance and to ensure it doesnt crash..?

]]> By: Xaprb http://www.xaprb.com/blog/2006/07/19/3-ways-to-maintain-rollup-tables-in-sql/#comment-1164 Xaprb Thu, 20 Jul 2006 16:44:51 +0000 http://www.xaprb.com/blog/?p=161#comment-1164 That’s one way I suppose, depending on the way the tables are built, but I can’t see that working well for my current employer’s tables. Finding orphans would involve an exclusion join in our schema, and could be much more involved depending on the schema — it might be much more difficult than zeroing or one of the other methods I mentioned. In our case I think it would be much harder on the DB server.

What would be the advantage of ON DUPLICATE KEY UPDATE over REPLACE, besides that it updates rows in place instead of deleting and inserting? (That’s a significant benefit anyway… I’ll look at our rollup queries again)

]]> By: groditi http://www.xaprb.com/blog/2006/07/19/3-ways-to-maintain-rollup-tables-in-sql/#comment-1158 groditi Thu, 20 Jul 2006 12:10:33 +0000 http://www.xaprb.com/blog/?p=161#comment-1158 Why not just use ON DUPLICATE KEY UPDATE and then run a delete for orphans at the end of the job?

]]>