Running the Knora API Server on a Production System

This section describes possible ways of running the Knora API server in an production environment. The description should only be taken as a first short introduction to this topic. Further reading of the referenced materials is advised.

Note

Our platform of choice is Linux CentOS 7 and is thus assumed in the description. The general idea should be usable on all platforms with small changes.

To run the Knora API server, we have two main components. First, the zipped distribution of the server and second a supported triplestore.

Creating and running the distribution package

Inside the knora/webapi folder run the following sbt commands:

` $ sbt packageBin `

This will create a zip file inside the knora/webapi/target/universal folder. To run the Knora API Server, unzip this package, and execute the webapi script inside the bin folder.

Alternatively, the command sbt stage will create a folder with the same content as before, but will skip the zipping step.

Running a supported triplestore

See the chapters on Starting Fuseki 3 and Starting GraphDB-SE on how to start a supported triplestore.