In my last two posts I talked a little bit about bitmap indexes, how they work and why they lock. I also mentioned that they should be used for low cardinality but that there is no cut-and-dried answer about what low cardinality is. I’m not starting here to talk about cardinality on bitmap indexes, actually Richard Foote did a much better job on that so if you interested in that check out the following blog posts from him:
- Myth: Bitmap Indexes with high distinct columns (Blow Out)
- Myth: Bitmap Indexes with high distinct columns (Supermassive Black Hole)
- So what is a good cardinality estimate for a bitmap index column?