The SUT component is responsible for executing the engine to be evaluated in a docker container. cAdvisor is also started, using the docker-compose.yml in this folder on the same machine as SUT engine, to monitor the resource usage of the engine's docker container.

In order to visualize the metrics, automatic dashboards for Grafana are also created by the create-dashboards.sh script. Check the monitoring unit component for more information about the underlying metrics collection using InfluxDB, and Grafana.

The scripts in this component are adapted from the RSPLab.

Running

The following command runs the specified engine with the given CLI args.
./start.sh -e [engine_folder] -- [CLI args]

The CLI args are dependent upon the underlying engines to be evaluated.

RMLStreamer-SISO

Consult the RMLStreamer-SISO repo to check the valid CLI args used by the RMLStreamer-SISO.

Example CLI args:

./start -e rml-streamer -- -c -m mapping.ttl -o output-folder

Dockerfile

In order to create engine-specific docker container easier, we provide template Dockerfile which could be used to create your own engine container for evaluation.

Job-submitter

This is the script which is responsible for starting the mapping process once the engine container is initialized. Significant changes to this script might be needed depending on how complex it is to start the underlying engine.