Running Fuseki with UI

Fuseki/UI can be run in a number of ways:

• As a standalone server
• As a service run by the operation system, for example, started when the machine boots
• As a Web Application inside a container such as Apache Tomcat or Jetty.

Fuseki is also packaged as a plain server “Fuseki Main” with no UI for use as a configurable SPARQL server, for building as a Docker container, and as a deployment and development standalone server. It supports the arguments used by the standalone server.

The configuration file is the same format for all forms of the Fuseki server.

See “Fuseki Configuration” for information on how to provide datasets and configure services using the configuration file.

Fuseki as a Standalone Server

This is running Fuseki from the command line.

To publish at http://host:3030/NAME:

where /NAME is the dataset publishing name at this server in URI space.

TDB2 database:

fuseki-server [--loc=DIR] [[--update] /NAME]

The --loc directory is either a TDB1 or TDB2 database. The directory DIR must exist. If the database in DIR does not exist, then a new database is created. By default this is a TDB2 database unless the argument –tdb1 is given.

An in-memory, non-peristent database (always updatable) is:

fuseki-server --mem /NAME

Load a file at start and provide it read-only:

fuseki-server --file=MyData.ttl /NAME

where MyData.ttl can be any RDF format, both triples or quads.

Administrative functions are only available from “localhost”.

See fuseki-server --help for details of more arguments.

Layout

When run from the command line, the server creates its work area in the directory named by environment variable FUSEKI_BASE. When run from the command line, this defaults to the current directory.

Fuseki layout

If you get the error message Can't find jarfile to run then you either need to put a copy of fuseki-server.jar in the current directory or set the environment variable FUSEKI_HOME to point to an unpacked Fuseki distribution.

Starting with no dataset and no configuration is possible. Datasets can be added from the admin UI to a running server.

Fuseki as a Service

Fuseki can run as an operating system service, started when the server machine boots. The script fuseki is a Linux init.d with the common secondary arguments of start and stop.

Process arguments are read from /etc/default/fuseki including FUSEKI_HOME and FUSEKI_BASE. FUSEKI_HOME should be the directory where the distribution was unpacked.

Fuseki as a Web Application

Fuseki can run from a WAR file.

The war file can be downloaded from the project downloads page. It should then be renamed to the webapp name “ROOT.war” (this means there no name in the URL) or “fuseki.war” (with a name /fuseki/) or some other choice of name.

FUSEKI_HOME is not applicable.

FUSEKI_BASE defaults to /etc/fuseki which must be a writeable directory. It is initialised the first time Fuseki runs, including a Apache Shiro security file but this is only intended as a starting point. It restricts use of the admin UI to the local machine.

When deploying as a web application a more fully featured Admin API is made available and described on the Fuseki Server Protocol (REST API) page.

Configuring logging

When running from a WAR file in a webapp container such as Apache Tomcat, the logging configuration comes from the file log4j2.properties in the root of the unpacked war file, e.g. /var/lib/tomcat9/webapps/fuseki/log4j2.properties.

The name of the file is taken from web.xml:

This only applies when running in a webapp container. When run from the command line, the server looks for log4j2.properties in the current directory and if not found, uses a built-in configuration.

This logging goes to the standard output.

Fuseki with Tomcat9 and systemd

systemd may be set to sandbox Tomcat9. The file area /etc/fuseki will not be writable to Fuseki. To enable this area, add ReadWritePaths=/etc/fuseki/ to the file /etc/systemd/system/tomcat9.service.d/override.conf, creating the file if necessary.

systemd also captures standard out and routes it to the system journal:

  journalctl -u tomcat9

To direct the output to the traditional location of /var/log/tomcat9/catalina.out use the StandardOutput setting in override.conf:

    [Service]
   
    # Allow access to the Fuseki area
    ReadWritePaths=/etc/fuseki/
    StandardOutput=file:/var/log/tomcat9/catalina.out