On 07/02/11 16:22, Lior Barnea wrote:
Hi all,
My name is Lior and I work at 3I-MIND.
We are checking several solutions for RDF framework and triple store DB.
I currently checking Jena with TDB and I have several questions :
1. When does the indexing take place ? is it when you execute
"model.read" line ?
If that's the way you load then data, then yes.
There are also a 2 bulkloaders.
2. Is there an option to load data (let's say triples) without indexing
and then schedule the indexing to later ?
No.
The bulkloader2, does phase the operations into load data, then index in
separate stages so if the issue if the bulk load phase that could be
adapted.
But contract of model.read is that the data is available for access
after the call returns and thsi is done by indexing incrementally.
3. Is there any monitor options for TDB ?
"monitoring" is a broad term.
There is some very basic JMX support.
There "--explain" functionality for monitoring query optimisation decisions.
Neither is very sophisticated.
Andy
Best regards,
Lior Barne'a
3I-MIND
