Comments on: Deleting from ARCHIVE tables http://ronaldbradford.com/blog/deleting-from-archive-tables-2008-07-10/ Expert times and information on MySQL Fri, 12 Mar 2010 22:45:29 -0400 http://wordpress.org/?v=2.8.4 hourly 1 By: Joe Izenman http://ronaldbradford.com/blog/deleting-from-archive-tables-2008-07-10/comment-page-1/#comment-381 Joe Izenman Wed, 09 Jul 2008 22:55:41 +0000 http://ronaldbradford.com/blog/?p=1094#comment-381 More in keeping with the intent of the Archive engine would just be to actually create different tables... have a permanently empty template table from which you could run CREATE TABLE LIKE statements, and then mark the date in the table name. Then you can safely just drop old tables at whatever interval you feel appropriate. More in keeping with the intent of the Archive engine would just be to actually create different tables… have a permanently empty template table from which you could run CREATE TABLE LIKE statements, and then mark the date in the table name. Then you can safely just drop old tables at whatever interval you feel appropriate.

]]> By: Brian Aker http://ronaldbradford.com/blog/deleting-from-archive-tables-2008-07-10/comment-page-1/#comment-380 Brian Aker Wed, 09 Jul 2008 19:08:12 +0000 http://ronaldbradford.com/blog/?p=1094#comment-380 Hi! I made Archive not delete data on purpose. The only way to delete data was via drop table (so a different set of privs). This was a feature request by Robin at the time. Personally I would call the above a bug, and I would suspect that having partitioning optimizing the delete to what must be a fast truncate means that it is not following the GENERATE macro. Which probably means that Partitioning is not handling HTON flags for different engines correctly :) Cheers, -Brian Hi!

I made Archive not delete data on purpose. The only way to delete data was via drop table (so a different set of privs). This was a feature request by Robin at the time.

Personally I would call the above a bug, and I would suspect that having partitioning optimizing the delete to what must be a fast truncate means that it is not following the GENERATE macro. Which probably means that Partitioning is not handling HTON flags for different engines correctly :)

Cheers,
-Brian

]]>