FaCTPlusPlusException - bad allocation

[GoogleCodeExporter]

I am not sure this is right place to ask, but I have problem with FaCT++.

I am writing Java program with OWLAI2.2.0. And I use FaCT++ as DL reasoner.

After starting program, it make exception, FaCTPlusPlusException - bad 
allocation.

Can you guide me what sorts of behaviors makes this "bad allocation"?

My program loads more than 100 ontologies (All is "Pizza" ontology, but 
have different logical ontology URIs)and alignment between them. It works 
during some time, but after that it makes errors.

Please help me. Thank you very much.

Original issue reported on code.google.com by dhkim1...@gmail.com on 15 Oct 2009 at 8:40

[GoogleCodeExporter]

This usually means that FaCT++ needs more memory for the reasoning than it is
available for it. So , you can treat is as a "Run out of memory" exception. 
However,
if you are sure that it is not a case, I'd like to see the ontology(-es) and 
what are
you doing with them to check what might it be.

PS. I assume you are using the latest version of FaCT++

Original comment by dmitry.t...@gmail.com on 15 Oct 2009 at 6:52

[GoogleCodeExporter]

Thanks very much for your comments. 

I am using FaCTpp-OWLAPI-v1.3.0.1.jar and FaCTpp-win-v1.3.0.zip on Windows XP 
pack3.

I run java with '-Xms 128m -Xmx1024m'. After runing my program, I monitored 
memory 
status with jvisualvm.exe included in JDK package(1.6.0_16). But, my program 
pops 
the exeception before reaching 1024m, specifically at the moment near 120m. 

About what am I doing,-
In constructor, I used following code, it called only one time during program 
life 
time - manager, reasonerFactory, reasoner are member objects of this class.
private OWLOntologyManager manager;
manager = OWLManager.createOWLOntologyManager();
private OWLReasonerFactory reasonerFactory;
reasonerFactory = new FaCTPlusPlusReasonerFactory();
private OWLReasoner reasoner; 
reasoner = reasonerFactory.createReasoner(manager);

And then, executes following pattern of code(called by other object), more than 
thousands of times.  
pattern: call block 2, whenever finishing block 1. means that at some time, 
reasoner 
always loads only 1~3 ontologies.

1{ 
ont = manager.reloadOntology(uri);
Set<OWLOntology> imports = manager.getImportsClosure(ont);
reasoner.loadOntologies(imports);
//some inferences ...(i.e. SubClassOf)
}

2{
reasoner.clearOntologies();
}

Attached files are some parameter of reloadOntology() function.
There are pizza0 ~ pizza99 ontologies. and pizza0-pizza1 includes alignments 
between 
pizza0 and pizza1. So, there are 100x100 alignments.

Execuse my English, Thanks you.

Original comment by dhkim1...@gmail.com on 16 Oct 2009 at 4:40

Attachments:

• pizza0.owl
• pizza1.owl
• pizza0-pizza1.owl

[GoogleCodeExporter]

1G for such a big problem is not enough. Increasing the memory by passing 
-Xmx4G or so will help.

Original comment by dmitry.t...@gmail.com on 8 Apr 2015 at 4:40

• Changed state: Done