Threading your table drops could be problematic if you are dealing with foreign key constraints. If constraints won't get in the way, just create a threaded program to execute multiple statements at a time. Are you just cleaning out a database? Is speed really a factor here? Please better describe your problem + environment.
-----Original Message----- From: E R [mailto:pc88m...@gmail.com] Sent: Wednesday, March 02, 2011 1:35 PM To: Dan Cutler Cc: dbi-users@perl.org Subject: Re: Suggest an approach for multi-statements? On Wed, Mar 2, 2011 at 2:22 PM, Dan Cutler <dcut...@dsm.net> wrote: > Hi all! > > I am seeking some advice. > > I need to be able to drop several thousand (more like .25 million) tables > from a database using DBI. I don't think ? place holders work in DROP TABLE statements. AFAIK, they only work in statements like INSERT, SELECT, UPDATE and DELETE, and then only to represent data values - not schema related names. I.e., I've never seen this work: UPDATE FOO SET ? = 1. Some suggestions: 1. Fork multiple processes to perform the drop table statements (i.e. use several database connections and processes to perform them in parallel.) 2. How about just start using a brand-new database schema? Then you'll start off with zero tables. Copy data from the original schema as needed. Then you can drop the 250 K tables at your leisure.
