Load Pre-computed Join

A tool has been created to load a pre-computed join. This tool will generate an index to support a pre-computed join on a user provided SPARQL query, and then register that query within Apache Rya.

Registering a pre-computed join

Generating a pre-computed join is done using Pig to execute a series of Map Reduce jobs. The index (pre-computed join) is associated with a user defined SPARQL query.

To execute the indexing tool, compile and run org.apache.rya.accumulo.pig.IndexWritingTool with the following seven input arguments: [hdfsSaveLocation] [sparqlFile] [instance] [cbzk] [user] [password] [rdfTablePrefix]

Options:

• hdfsSaveLocation: a working directory on hdfs for storing interim results
• sparqlFile: the query to generate a precomputed join for
• instance: the accumulo instance name
• cbzk: the accumulo zookeeper name
• user: the accumulo username
• password: the accumulo password for the supplied user
• rdfTablePrefix : The tables (spo, po, osp) are prefixed with this qualifier. The tables become: (rdf.tablePrefix)spo,(rdf.tablePrefix)po,(rdf.tablePrefix)osp

Using a Pre-computed Join

An example of using a pre-computed join can be referenced in org.apache.rya.indexing.external.ExternalSailExample