Fuseki

Table of Contents VoID and SPARQL Service Description Documentation Staging SPARQL Endpoint URI How to start Fuseki in update mode Importing data using SPARQL 1.1 Graph Store HTTP Protocol How to import (put) RDF into Fuseki's store How to import (post) RDF into Fuseki's store Using GraphPusher tool Importing directly to TDB storage Optional help Production How to start Fuseki Synching Staging and Production Synch

VoID and SPARQL Service Description Documentation

VoID + SD: http://data-gov.ie/void.ttl

Staging

The TDB assembler file mentioned below sets the store location. For staging, it is located in /usr/lib/fuseki/DBStaging/

SPARQL Endpoint URI

http://staging.data-gov.ie/sparql

How to start Fuseki in update mode

Note that we start fuseki in read and write mode:

/usr/lib/fuseki/fuseki-server.datagovie.staging --verbose --update --port=3131 --desc

/usr/lib/fuseki/tdb2.datagovie.staging.ttl /dataset

Importing data using SPARQL 1.1 Graph Store HTTP Protocol

How to import (put) RDF into Fuseki's store

/usr/lib/fuseki/./s-put --verbose http://localhost:3131/dataset/data http://data-gov.ie/graph/{name} file.rdf

Note: Put clears and deletes existing graph and its data.

How to import (post) RDF into Fuseki's store

/usr/lib/fuseki/./s-post --verbose http://localhost:3131/dataset/data http://data-gov.ie/graph/{name} file.rdf 

Note: Post will add triples to existing graph. If graph doesn't exist, it will create one.

Using GraphPusher tool

URL: https://github.com/csarven/graphpusher Examples:

ruby GraphPusher.rb http://example.org/void.ttl --assembler=/usr/lib/fuseki/tdb2.datagovie.staging.ttl

ruby GraphPusher.rb http://example.org/void.ttl --dataset=http://localhost:3131/dataset/data

(See README for more information)

Importing directly to TDB storage

Example:

java tdb.tdbloader --desc=/usr/lib/fuseki/tdb2.datagovie.staging.ttl --graph=http://purl.org/linked-data/cube cube.ttl

Optional help

• Run the following bash script from a directory with several RDF files which needs to go into the same graph.

Remove comments and use accordingly for your needs:

#PRWODI=$PWD

#BN=$BASENAME

pwdrelative=${PWD##*/}

for file in *.ttl;

    do

#        filename=$(basename $file);

#        extension=${filename##*.};

#        graph=${filename%.*};

        /usr/lib/fuseki/./s-post --verbose http://localhost:3131/dataset/data http://data-gov.ie/graph/$pwdrelative $file;

    done;

Production

How to start Fuseki

Note that we start it in read only mode:

/usr/lib/fuseki/fuseki-server --verbose --desc /usr/lib/fuseki/tdb2.dataset.ttl /dataset

Synching Staging and Production

1. Review data in Staging.
2. Go to step 1.

Synch

NOTE: I have a graphpusher-hotswap.rb in works. I'll put that here when ready for production use. -Sarven

We do DB hot-swap:

1. Stop the production Fuseki service
2. Just caution: Make sure staging store exists (i.e., /usr/lib/fuseki/DBStaging ) and contains data. Double check staging.
3. Back up existing production store: mv DB DB.bak
4. Stop the staging Fuseki service (in case something funky updates are happening)
5. Make a copy of the staging store for production: cp -R DBStaging DB
6. Start production Fuseki service
7. Start staging Fuseki service
8. Check around http://data-gov.ie/ to make sure things are running okay.
9. Go outside and get some sun