Re: [Virtuoso-users] Multiple core

Hi Marc, A single query runs on single thread (plus I/O in separate thread) in Virtuoso. Without waiting for disk, it's faster to run it in a single thread than to parallelize and get synchronization overheads (and bug-prone complications). With waiting for disk, CPU times does not matter -- i

Re: [Virtuoso-users] Multiple core

Hi Marc, On Tue, Jun 10, 2008 at 11:48 PM, Marc-Alexandre Nolin wrote: > Hello, > > How can I compile or fine tune a Virtuoso installation to take > advantage of all the core I have in the server. I currently have a 4 > core server and when I query it, in the top utility on Linux, only 1 > co

[Virtuoso-users] Multiple core

Hello, How can I compile or fine tune a Virtuoso installation to take advantage of all the core I have in the server. I currently have a 4 core server and when I query it, in the top utility on Linux, only 1 core is having a load on him. I'm about to try Virtuoso on a 16 core and a 32 core in the

Re: [Virtuoso-users] performance tuning

Hi Hugh, Indeed, now it works. Here you are the first report: Query result: REPORT VARCHAR { Precode: 0: $25 "OBO#GO_0005737" := Call DB.DBA.RDF_MAKE_IID_OF_QNAME_SAFE ((http://www.semantic-systems-biology.org/ontology/rdf/OBO#GO_0005737)>) 7: $26 "ema#label" := Call DB.DBA.RDF_MAKE_IID_OF_Q

Re: [Virtuoso-users] performance tuning

Hi Erick, Sorry, I forgot to have you prefix the queries with the keyword 'sparql' to tell Virtuoso to execute via the SPARQL rather than SQL engine, so the query below would be: explain('sparql prefix rdf: select distinct ?ontology from