On 5/22/14 8:33 AM, Bart Vandewoestyne wrote:
On 2014-05-18 16:11, Hugh Williams wrote:>Hi Sumit > >Ah, didn't realise you started with a v6 row store database, thus how >did you upgrade to v7 ? > >As if you simply check pointed and restarted the v6 database with a v7 >binary then it would be automatically upgraded but all the tables >including the RDF_QUAD tables and indexes would remain in the original >row store mode. > >There are two options for converting the RDF_QUAD table to column store >which are: > >1. To dump the current RDF triples to datasets and reload them in an >empty database which will automatically then create the RDF_QUAD tables >and indexes in column wise mode. Scripts for dumping and reloading the >RDF data can be found at: > >http://virtuoso.openlinksw.com/dataspace/doc/dav/wiki/Main/VirtRDFDatasetDumpHugh and list,I just wanted to report my findings with this first option: I have a virtuoso.db file of 142 GB and it took one night to dump it to .ttl and .graph files. The .ttl and .graph files together take 242 GB. Re-loading the .ttl files into the RDF triple store seems to take more time... I'm loading for more than 24 hours now, and I'm at about 60 % of the total tuple count. Is it normal that re-loading the data takes significantly more time than dumping it? Regards, Bart
Yes, for the following reasons: 1. Exporting is just about writing to files2. Importing is about indexing and writing to internal identifiers (named graph IRIs).
You can speed up you loads by running multiple loaders in parallel. Even better if you are using the cluster edition as the same source files could be copied to multiple physical cluster nodes where each is running multiple loaders.
Links:[1] http://bit.ly/11pPq6F -- old bulking load report re., Virtuoso 7 and LOD Cloud Cache.
-- Regards, Kingsley Idehen Founder & CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog: http://www.openlinksw.com/blog/~kidehen Twitter Profile: https://twitter.com/kidehen Google+ Profile: https://plus.google.com/+KingsleyIdehen/about LinkedIn Profile: http://www.linkedin.com/in/kidehen
smime.p7s
Description: S/MIME Cryptographic Signature
------------------------------------------------------------------------------ "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE Instantly run your Selenium tests across 300+ browser/OS combos. Get unparalleled scalability from the best Selenium testing platform available Simple to use. Nothing to install. Get started now for free." http://p.sf.net/sfu/SauceLabs
_______________________________________________ Virtuoso-users mailing list Virtuoso-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/virtuoso-users
