Whilst testing a Solr install for a future MySQL -> Solr migration, it’s immediately apparent that the “rules” for what constitutes a good data stucture, and by extension an efficient search, are very different in Solr when compared to an RDBMS like MySQL. The most obvious thing being that data isn’t (or doesn’t seem to be) normalised to the same degree.
Does anyone have any advice regarding the best way to go about making the transition from MySQL to Solr? Are there any established patterns for structuring data in a non-RDBMS (Solr specifically) that I should be learning about? Any common pitfalls to avoid? Is it simply a case of de-normalising related tables into objects?
First of all, you have to ask yourself if you want to:
For anything other than non-trivial relational schemas, I’d recommend #2. The more heterogeneous data you have in one index, the less useful it is.