Folks:
On my projects, I haven't found PostgreSQL's implementation of clustered
indexes to be particularly useful ... gains of only a few percent in query
efficiency in exchange for a substantial management task. Obviously, not
everyone has had the same experience, or we wouldn't still have the feature.
When I re-vamp my articles on indexing, I would like to include something
about clustered indexes. Can people give me some examples of cases where
they have found clustered indexes to be useful, preferably with some
statistics?
--
-Josh Berkus
Aglio Database Solutions
San Francisco