I see, seems that in my case the fst have to work particularly hard since the number of unique terms is huge, and I use * in both parts of the query
Thanks for the detailed answer! -- Sent from: http://lucene.472066.n3.nabble.com/Lucene-General-f642108.html
