owlim-discussion  

Re: [Owlim-discussion] Problem with RepositoryResult

Damyan Ognyanoff
Wed, 01 Sep 2010 01:59:04 -0700

Hi Tobias,

Thank you for the feedback. Indeed, there is a such fixed timeout (5000ms) for 
that kind of requests, but those are enforced only for when "repository-type" 
config parameter is set to "weighted-file-repository" e.g. RDFRank and 
spreading activation components are enabled and used. This approach was chosen 
since in the past, our public demo services were fully exposed through a Sesame 
REST full API, which one very easily could made them unusable, if such internal 
cut-off was not present.

It is now a bit obsolete since we currently use our own frontend for such 
purposes so that leftover will be fixed/removed/made configurable in the next 
release (planned to get out somewhere in the mid October). As a workaround you 
could make use of a suitable sparql query to fetch such data out of your 
repository. Such time-out during suery evaluation is still present but it is 
enabled and configurable through "query-timeout" config parameter/JVM runtime 
property as stated in the sysdoc.

HTH,
Damyan Ognyanov
Ontotext AD.
  ----- Original Message ----- 
  From: Tobias Mathäß 
  To: owlim-discussion@ontotext.com 
  Sent: Wednesday, September 01, 2010 10:55 AM
  Subject: [Owlim-discussion] Problem with RepositoryResult


  Hello,

   

  I am currently working with a bigowlim-rdf-store, and I am having some 
difficulties with large query results. What I want to do is initially iterate 
over all statements in the store in order to extract some statistical data 
before starting my application. The problem is, that the 
RepositoryResult.hasNext() method returns false exactly 5 seconds after the 
result is created with the call.

   

  RepositoryResult<Statement> res = con.getStatements(null, null, null, false);

   

  This problem also occurs, if I try to work around this problem by calling 
res.asList() and iterating over the list of statements (with this approach I 
get more results then by iterating directly over the store since the 
transcription of the Statements to the list is much faster than the actual 
iteration and analysis of the data, but for very large datasets, the 5 seconds 
are not enough to transcribe all statements in the list)

   

  no matter how many statements are actually in the store. This is by far not 
enough time to iterate the whole result set.

   

  Is there any configuration-parameter I can manipulate in order to keep the 
resultset open for more then five seconds? I read the documentation, but did 
not find anything helpful so far.

   

  I am using bigowlim 3.3, together with the latest version of sesame (2.3.2) 
and Lucene 3.0.0.

   

  It would be great if you could help me with this problem.

   

  Regards,

  Tobias Mathaess

   

   



------------------------------------------------------------------------------


  _______________________________________________
  OWLIM-discussion mailing list
  OWLIM-discussion@ontotext.com
  http://ontotext.com/mailman/listinfo/owlim-discussion
_______________________________________________
OWLIM-discussion mailing list
OWLIM-discussion@ontotext.com
http://ontotext.com/mailman/listinfo/owlim-discussion