[OSM-dev] Too many slow queries in db

Richard Fairhurst richard at systemeD.net
Tue Sep 4 09:17:01 BST 2007

Brett Henderson wrote:

> Can we build a list of possible improvement areas over current
> database?  This would less any alternatives be measured quantitively.
> For example:
> * The current top 10 slow queries.  Assuming an alternative schema still
> follows the nodes, segments, ways model this would let us manually run
> queries against alternative implementations without porting tools.

I anticipate we'll get a measurable speedup once 0.5 goes live.  
Certainly, Potlatch's queries will be greatly improved by no longer  
having to join everything to the segment tables (and some of them can  
be eliminated entirely).


More information about the dev mailing list