Virtuoso performance

Profiling

http://vos.openlinksw.com/owiki/wiki/VOS/VOSBitmapIndexing

http://vos.openlinksw.com/owiki/wiki/VOS/VirtFixBrokenQuadStoreIndex

http://vos.openlinksw.com/owiki/wiki/VOS/VirtRDFPerformanceTuning

http://docs.openlinksw.com/virtuoso/ptune/

http://docs.openlinksw.com/virtuoso/perfdiag/

http://docs.openlinksw.com/virtuoso/fn_prof_enable/

http://vos.openlinksw.com/owiki/wiki/VOS/VirtSPARQLPrefOrder

output:format


Answered on the Virtuoso Users mailing list where the question was also asked... Dumping results in various formats can be done by using the define output:format "{XX}" pragma, so in your case it would be: SQL> sparql define output:format "TURTLE" CONSTRUCT ... Other possible formats are: NICE_TTL RDF_XML etc. When using the ISQL client to fetch long texts, use the set blobs on; directive to avoid receiving a data truncated warning. i.e.: SQL> set blobs on; SQL> sparql define output:format ... For CONSTRUCT, the supported formats are: TRIG, TTL, JSON, JSON;TALIS, SOAP, RDF/XML, NT, RDFA;XHTML, JSON;RES, HTML;MICRODATA, HTML, JS, ATOM;XML, JSON;ODATA, XML, CXML;QRCODE, CXML, HTML;UL, HTML;TR, JSON;LD, CSV, TSV, NICE_TTL, HTML;NICE_MICRODATA, HTML;SCRIPT_LD_JSON, HTML;SCRIPT_TTL, HTML;NICE_TTL

iSQL

isql 1111 errors=stdout <test.sql >test.out
isql 1111 dba dba VERBOSE=OFF 'EXEC=status()' test.sql test2.sql -i arg1 arg2

http://docs.openlinksw.com/virtuoso/isqlgentesttool/ http://docs.openlinksw.com/virtuoso/invokingisql/

SPARQL query limiting results to 100000 triples

https://community.openlinksw.com/t/sparql-query-limiting-results-to-100000-triples/2131

Prevent Limits of Sorted LIMIT/OFFSET query

SELECT ?p ?s
WHERE
  {
    {
      SELECT DISTINCT ?p ?s
      FROM <http://dbpedia.org>
      WHERE
        {
          ?s ?p <http://dbpedia.org/resource/Germany>
        } ORDER BY ASC(?p)
    }
  }
OFFSET 50000
LIMIT 1000

http://docs.openlinksw.com/virtuoso/rdfsparqlimplementationextent/

RAID 10 vs Virtuoso striping. What is the best for IO performance?

https://community.openlinksw.com/t/raid-10-vs-virtuoso-striping-what-is-the-best-for-io-performance/1407

Backward chaining

https://sourceforge.net/p/virtuoso/mailman/message/34144456/

Stripping

https://sourceforge.net/p/virtuoso/mailman/virtuoso-users/?viewmonth=201003

Efficient Use of SQL - SQL Execution profiling

profile ('SELECT COUNT (*) from ct_c a, ct_c b WHERE a.k2 = b.k1 option (loop, order)');

Recommendations

CPU load goes up to 100% when Virtuoso runs its look up processing

Virtuoso.ini param

How to Update Large SPARQL Data avoiding due to database checkpoint abortion?

Translate and Analyze modes for analyzing sparql queries

The “MaxQueryCostEstimationTime"param should never be set in the INI file

Compiling and configuring Virtuoso

Upgrading to the latest Virtuoso git develop/7 branch which is a 07.20.3219

Returns statistics for a running server as a result set

status('rhck');

Virtuoso 7 typically requires 10GB RAM per billion triples,

What is the storage cost per triple?

Virtuoso.ini settings parameters

Performance diagnostics

http://docs.openlinksw.com/virtuoso/perfdiag/

IO Bound

Configuring Virtuoso for Scale

Performance Tuning Virtuoso for RDF Queries by using RDF indexes