I have recently switched the back end of my Fuseki2 server from TDB to TDB2.
In the past I have used the tdbstats command line tool for TDB to generate a statistics file, to help the query optimizer, but there is apparently no equivalent for TDB2. My TDB2 dataset is built using the tdb2.tdbloader utilityl, from an N-Quads file of 3.5M triples. I have some CONSTRUCT queries which run rather slowly, and I wondered if TDB2 needed a statistics file as TDB used to, or if there's some other optimization step I might be missing. -- Conal Tuohy http://conaltuohy.com/ @conal_tuohy +61-466-324297
