In 2.15, we added the wildcard field type that works well for cases where you want arbitrary substring matches (versus full-text search).
While we added unit tests and integration tests, we did not do any benchmarking. In particular, it would be nice to see how matches on arbitrary substrings on a wildcard field compare to similar queries on text/keyword field types. (Also, it would be interesting to see how much worse exact matches perform on wildcard fields versus text/keyword.)
We should be able to add wildcard fields to the http_logs and Big5 workloads. (It's extra indexing work, so we need to keep in mind that indexing numbers will get worse.)
After adding wildcard operations to http_logs and Big5, we could also create a separate test procedure, that is limited to text / keyword queries, for users to easily benchmark and compare the results.
In 2.15, we added the
wildcard
field type that works well for cases where you want arbitrary substring matches (versus full-text search).While we added unit tests and integration tests, we did not do any benchmarking. In particular, it would be nice to see how matches on arbitrary substrings on a wildcard field compare to similar queries on text/keyword field types. (Also, it would be interesting to see how much worse exact matches perform on wildcard fields versus text/keyword.)
We should be able to add wildcard fields to the http_logs and Big5 workloads. (It's extra indexing work, so we need to keep in mind that indexing numbers will get worse.)