First off, I'm going to start this posting with a description of how Drupal works traditionally and how BigCloset works now. Some of this information is a copy and paste of a Comment I made on a blog.

Traditionally everything in Drupal is stored in a MySQL or similar database.

These constant and continual use of MySQL for everything causes us much trouble when it comes to the content of this site. Nearly everything is stored in MySQL and so almost all data needs to be "retrieved" from MySQL including the stories. In a large enterprise situation this is offset by having several MySQL servers and doing load balancing, but right now, that's not in the budget for us.

One of the reasons we were able to open search all the time to everyone is we "off-loaded" search to a separate server which uses it's own database system and provides search facilities and all kinds of ability to "refine your search" using what we call "Facets" or "Faceted Search".

One of the features of this new system is that we can tell the search server, to "intercept" all "Taxonomy Links" in hopes of reducing our database load.

It doesn't look quite so pretty now, but in the coming days/weeks we will be working on changing things so that you will get proper "teaser" blocks for all the stories returned by your search.