Hi Dave,
Just wanted to piggyback on Rich's comments with regards to secondary indices.
A secondary indexing capability with a cost-based query planner is actively being developed as part of our next release.
The initial capability will provide the ability to configure secondary indices at the attribute-level to support the following types of CQL query predicates:
- exact match of text, numeric, or date/time attributes
- comparison operations for numeric attributes
- comparison operations for temporal attributes
- range operations for numeric and temporal attributes
The very large 2ND_IDX_NGRAM_2_4 table that are you seeing for text fields was aimed towards supporting fuzzy text matching by indexing n-grams (bigrams and trigrams) for text fields, however this is an approach that are are moving away from (at least for the initial capability). 30 million entries for 150 thousand records does not really surprise me given the current implementation, however it is currently not wired to any sort of query planning capability for you to take advantage of it. Stay tuned and hopefully we will have something that suits your needs in the near future.
Please let me know if you have any questions.