[
https://issues.apache.org/jira/browse/MARMOTTA-492?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13978888#comment-13978888
]
Sebastian Schaffert commented on MARMOTTA-492:
----------------------------------------------
Hi Maxime,
by default the KiWi triple store does not itself optimize SPARQL queries and
instead relies on the default Sesame approach, which computes the JOIN
in-memory, and therefore, as you say, produces 60000 * 60000 results before
filtering in-memory. However, KiWi comes with optional SPARQL-SQL translation
which should give you optimized performance. To use it, you have to add the
kiwi-sparql artifact to your classpath and wrap the KiWiStore in a
KiWiSparqlSail as follows:
Sail store = new KiWiSparqlSail(new KiWiStore(conf));
Can you try this and tell us if it improves your performance?
> Issue With a SPARQL Query
> -------------------------
>
> Key: MARMOTTA-492
> URL: https://issues.apache.org/jira/browse/MARMOTTA-492
> Project: Marmotta
> Issue Type: Improvement
> Components: KiWi Triple Store
> Affects Versions: 3.2.0
> Reporter: Maxime Poitevineau-Millin
> Assignee: Sebastian Schaffert
> Labels: SPARQL, optimization, postgresql
>
> We are currently implementing marmotta in our system but!
> When trying this query :
> {code}
> SELECT * FROM <sesame:nil> {
> ?kw rdf:type Keyword .
> ?kw2 rdf:type Keyword .
> ?kw2 ?rel ?kw .
> }
> {code}
> on a database containing 60 000 keywords, marmotta is creating 60 000 * 60
> 000 queries similar to this one :
> {code:sql}
> SELECT
> id,subject,predicate,object,context,deleted,inferred,creator,createdAt,deletedAt
> FROM triples WHERE deleted = false AND subject = '458172920399523841' AND
> object = '458172732532453376' AND context = '458172711103754240'
> {code}
> Which obviously takes a long time.
> This query takes 2 sec on Virtuoso or Owlim, is there any configuration issue
> or something we should change?
--
This message was sent by Atlassian JIRA
(v6.2#6252)
