Egon, On 15 Oct 2013, at 22:35, Egon Willighagen <egon.willigha...@gmail.com> wrote:
> Dear Hugh, > > On Sat, Oct 12, 2013 at 3:36 PM, Hugh Williams <hwilli...@openlinksw.com> > wrote: >> Glad to hear the stability has improved ... > > OK, it had crashed again in the past half day ... > >> With regards to SNORQL, what happens if you run the same query directly >> against the Virtuoso SPARQL endpoint ie /sparql rather than /snorql ? As >> SNORQL is also installed on the DBpedia SPARQL endpoint we host and also >> takes time running against /snorql but not against /sparql . It also >> takes time to run against a native D2R /snorql service like >> http://dblp.l3s.de/d2r/snorql , so this seems to be an overhead of SNORQL ? > > I tried that, but it times out on the SPARQL end point too... in both > cases it timed out on: > > SELECT DISTINCT ?property > WHERE { [] ?property [] } > ORDER BY ?property [Hugh] Running this query against the DBpedia /sparql endpoint it returns in about 20secs, whereas running against its /snorql interface it hasn't returned in the past 15mins and will probably timeout. What is he current triple count of your Virtuoso Server instance. Also when the database is "warm" can you please provide the output of running the "status();" command from isql to see the state of the database. > > Just to remove some possible source of more time needed, I removed the > ORDER BY... then it runs fine in both cases! > > That surprises me, because there are only 50 ?property values... it > sounds as if the ordering is not done on the result list, but before > the applying the DISTINCT? Is that possible? Seriously, I have no > knowledge of how tools actually perform a query, and maybe this SPARQL > is just not right... > > Anyway, I removed the ORDER BY from the SNORQL code, so that when > people click "Properties" they get an unsorted list, but at least not > a time out... getting all properties is normally one of the first > things I run on a SPARQL end point, just to get a feeling of the data > that is there... the classes won't help in my case, because each and > every chemical structure is modeled as a class... > > Thanks for your pointers! I'm happy that this works. I'm hoping this > will also reduce the stress on the system a bit... still, with a > public SPARQL end point, I'm impressed to see VOS behave so stable, > even though VOS6 was (I had to restart that really just a few times on > the course of two years) more stable... [Hugh] We will be able to get Virtuoso 7 to the same level of stability ... Regards Hugh > > Egon > > -- > Dr E.L. Willighagen > Postdoctoral Researcher > Department of Bioinformatics - BiGCaT > Maastricht University (http://www.bigcat.unimaas.nl/) > Homepage: http://egonw.github.com/ > LinkedIn: http://se.linkedin.com/in/egonw > Blog: http://chem-bla-ics.blogspot.com/ > PubList: http://www.citeulike.org/user/egonw/tag/papers > ORCID: 0000-0001-7542-0286 > > ------------------------------------------------------------------------------ > October Webinars: Code for Performance > Free Intel webinars can help you accelerate application performance. > Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from > the latest Intel processors and coprocessors. See abstracts and register > > http://pubads.g.doubleclick.net/gampad/clk?id=60135031&iu=/4140/ostg.clktrk > _______________________________________________ > Virtuoso-users mailing list > Virtuoso-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/virtuoso-users
smime.p7s
Description: S/MIME cryptographic signature
------------------------------------------------------------------------------ October Webinars: Code for Performance Free Intel webinars can help you accelerate application performance. Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from the latest Intel processors and coprocessors. See abstracts and register > http://pubads.g.doubleclick.net/gampad/clk?id=60135031&iu=/4140/ostg.clktrk
_______________________________________________ Virtuoso-users mailing list Virtuoso-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/virtuoso-users
