You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 4 Next »

Overview

A publicly-accessible SPARQL endpoint may be installed.  It takes the form of an additional repository which contains everything in the main repository which is public (i.e. not email-addresses, phone-numbers, etc.; not un-published resources).  It is affectionately known as the "sparqler" associated with this main repository.  It is synchronized with its main repository on a regular basis - typically every night.

Installation

The installation of a sparqler follows very much the same procedure as for the main repository, with one or two changes because it doesn't live at the same location.  The quick-start installation procedure is as follows - additional details are the same as for the main repository.

  1. In the existing eagle-i home directory (the parent of the repository home-directory, ${REPO_HOME}) create a directory "sparqler", and set an environmental variable ${SPARQLER_HOME}.   For example, if the repository home is /opt/eaglei/repo, then the sparqler home would be /opt/eaglei/sparqler.
  2. Copy the file ${REPO_HOME}/webapps/sparqler.war to Tomcat's webapps directory.
  3. Add an additional system property to catalina.properties:
org.eaglei.sparqler.home=/opt/eaglei/sparqler
  1. Run the script ${REPO_HOME}/etc/prepare-install.sh with an additional argument: "sparqler-users.derby"
  2. Create and edit the repository configuration file (${SPARQLER_HOME}/configuration.properties) - it should be identical to the main repository's config-file, with the following two exceptions - we set: 
    eaglei.repository.sesame.dir=${sys:org.eaglei.sparqler.home}/sesame
eaglei.repository.log.dir=${sys:org.eaglei.sparqler.home}/logs
  3. Start up Tomcat
  4. Run the script ${REPO_HOME}/etc/finish-install.sh, using the username and password that you want for administering the sparqler (this should be different from the main repository's superuser).

Synchronization

The sparqler has a synchronization program,  synchronize-sparqler.sh (found in folder ${REPO_HOME}/etc), that should be run (as the same user that runs tomcat) every night.  On a Unix-like system (including OS X and above on a Mac), the system administrator should set a line in a crontab file to accomplish this (something like the following):

13 2 * * *    tomcat   cd /opt/eaglei && bash repo/etc/synchronize-sparqler.sh

(the first 6 fields are "min", "hr", "day-of-month", "day-of-week", "user";  "*" is a wild-card - so this would run daily at 02:13am as the user "tomcat").  

If you run on Windows (warning: this is at your own risk - we have not tested the system on Windows), you should set up a Unix-like environment (e.g. Cygwin) in which to perform the installation and run the shell-scripts.

Queries

See also eagle-i SPARQL Queries for Monitoring and Maintenance, which contains a library of pre-written SPARQL queries to try.

  • No labels