> Does somebody ever load something has large has a complete N3 version > of GenBank or Refseq into a single Virtuoso Triplestore? I'm using a > the ttlp_mt program has mentioned on how to load Bio2RDF data, but I > call it from a Perl script.
We've loaded B3S dataset plus some extra data without performance troubles, so I see no reason to worry in your case. > Is there a way to enable compression of the object part of the triple if its > a literal because > I've plenty of sequence in these dump that take a lot of space when > not compressed? There's no special compression of objects, but version 6 compacts database pages so there's little reason for extra compression anyway. > Also, is there a faster way to load than ttlp_mt for > N3 because the load is slowing down? While the database is in personal use and consists only data from outer sources, you may wish to disable logging while loading. More important, set appropriate number of buffers and, if possible, make the database striped. Refer to http://docs.openlinksw.com/virtuoso/dbadm.html, esp. to section 6.1.9.1.15. [Striping] for more details. Best Regards, Ivan Mikhailov OpenLink Software http://virtuoso.openlinksw.com
