Re: Index Scans become Seq Scans after VACUUM ANALYSE

Michael Loftis <mloftis@wgops.com>

From: Michael Loftis <mloftis@wgops.com>
To: pgsql-hackers@postgresql.org
Date: 2002-04-17T09:17:50Z
Lists: pgsql-hackers
My opinion.

Expose some of the cost factors via run-time settings (or start-time 
settings).

This would allow those who wanted to 'tweak' the planner to do so and 
those that felt the defaults were fine or didn't know to leave them alone.

Comments?