Hardware
Below we list the hardware specifications of the two basic hosts for developing and perform basic benchmark testing with the GeoRDFBench Framework.
Development Host
The development platform for GeoRDFBench is a VirtualBox machine, host name ubuntu-vma-tioa, operating system Ubuntu 18.04.6 LTS, 11GB RAM, 4 vCore CPU Intel i7 class and 80GB filesystem. The host configuration in the JSON Specification Library is ubuntu_vma_tioaHOSToriginal. Apart from development, this configuration is sufficient for test benchmarks with datasets of up to 10M triples.
Baseline Benchmarking Host
The hardware platform is an Intel NUC8i7BEH box, host name NUC8i7BEH, operating system Ubuntu 22.04.2 LTS, with 32GB DDR4-2400MHz, 8 executing threads=4 cores * 2 threads/core Intel i7 class, a Samsung SSD NVMe 970 EVO Plus 500GB system disk and a secondary data disk Western Digital WDC WD20SPZX-75U 2TB mounted on /data. The host configuration in the JSON Specification Library is nuc8i7behHOSToriginal. This configuration is sufficient for benchmarking datasets of up to 100M triples. The main bottleneck is the data disk which is the 2.5" 5400rpm HDD.
Software
All Hosts
The suggested operating system is Ubuntu >= 18.04.6 LTS with JDK-8 installed. Some RDF modules such as GraphDBSUT, StradogSUT and VirtuosoSUT require that the corresponding system has been installed in the host. The versions that must be installed, after explicitly requesting them or the license key from the software vendors, are: (i) graphdb-free-9.11.1-dist.zip, (ii) stardog-latest7.9.1.zip, (iii) virtuoso-opensource-7.2.9.tar.gz. Additionally, a PostgreSQL v14 + PostGIS v3 installation is required for running Strabon. The PostgreSQL installation also serves as the custom report sink where experiment results are recorded. For experiments not involving Strabon, PostgreSQL as a custom report sink may very well be served by an existing remote installation.
Finally, the GeoRDFBench Framework is required. The pre-built binary+source code can be downloaded from the following link geordfbench-v1.1.0.zip
while the latest version of the source code can be found in our public
Github repository https://github.com/tioannid/geordfbench. Detailed instructions for both installation options are provided in this site.
Development Host
For development purposes it is better to use the Github oriented installation method. In this case, in addition to the software requirements for all hosts, mentioned earlier, it is also necessary to have, an IDE (Integrated Development Environment) such as Netbeans >=8.2, the build tool Apache Maven >= v3.6.3 and Git >= v2.17.1 as the source code version control system.