On 09/08/13 21:10, Martin Vassilev wrote:
On 08/09/2013 06:48 PM, Andy Seaborne wrote:
On 09/08/13 16:42, Martin Vasilev wrote:
Hi,
Just installed Jena 2.10.1 and playing around with the tdbloader,
tdbloader2 and tdbquery cmd tools. The strange thing is when I execute
tdbquery I get java.lang.OutOfMemoryError: Java heap space exception.
Doing the following commands:
./tdbloader --loc=../temp/
../../edubase-rdf-r2rc3/dataset/EduBaseExtract.rdf
The EduBaseExtract.rdf has about 6.2 million triples. The triples are
loaded without errors - Then I execute the following :
./tdbquery --loc=../temp/ --query query.txt --time -v
I read jvm may be limited for 1.5 or 2 GB of RAM but should the amount
of ram be bound to the size of the store? I mean if that is the problem
and 2 GB ram are not enoght for 6.2 million how much ram a system has to
have for a 60 or 600 mil. store?
It's usually enough.
Regards,
Martin
What's in query.txt?
Andy
Hi, There is just " select * where {?s ?p ?o} ". Where is the value
defining the max amount of ram used by jena ? Maybe it needs to be
increased...
Martin
./tdbquery --loc=../temp/ --query query.txt --time -v
You have chosen the default output format, which is formatter plain
text. To do that, the text formatter needs to calculate column widths
and to do that it needs to scan all the results, go back then output the
text now it knows the padding to add to columns.
That needs an in-memory copy for the intermediate result set.
Choose a format such as TSV or JSON that is streaming. --results=tsv or
--results=json.
Andy
