On 20/07/12 19:03, Andy Seaborne wrote: > Paolo has in the past looked at MapReduce jobs for very large scale > loading. Paolo?
Code is here: https://github.com/castagna/tdbloader4 ... not actively maintained at the moment (and building B+Trees isn't a great fit for MapReduce). However, it should work (and it would make a lot of sense if you already have an Hadoop cluster available). If you do not have an Hadoop cluster, then tdbloader2|3 are a best option. (I still need to check if tdbloader4 is affected by the same bug of tdbloader3 which Andy fixed recently... probably yes). tdbloader2 with enough RAM should work just fine for DBPedia (as Andy said). Olivier, were you using tdbloader or tdbloader2? Paolo
