Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/mchaloupka/r2rml-benchmark
Scripting to perform a benchmark of virtual SPARQL endpoints based on R2RML mappings
https://github.com/mchaloupka/r2rml-benchmark
Last synced: 23 days ago
JSON representation
Scripting to perform a benchmark of virtual SPARQL endpoints based on R2RML mappings
- Host: GitHub
- URL: https://github.com/mchaloupka/r2rml-benchmark
- Owner: mchaloupka
- License: mit
- Created: 2020-01-04T20:34:23.000Z (about 5 years ago)
- Default Branch: master
- Last Pushed: 2024-02-06T09:29:17.000Z (11 months ago)
- Last Synced: 2024-12-06T21:52:58.777Z (29 days ago)
- Language: F#
- Size: 3.52 MB
- Stars: 1
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# r2rml-benchmark
Scripting to perform a benchmark of virtual SPARQL endpoints based on R2RML mappings.
## Prerequisities
* Docker
* .NET Core SDK 5 or newer
Has been currently tested only on Windows and Linux. Should be easy to add support for other platforms, please create a pull request if it does not work on other platforms.
## Running the benchmark
To run the benchmark, run the following commands:
```
dotnet fsi main.fsx
```
It is possible to run it with different configurations:
* `dotnet fsi main.fsx quick` - to run a quick variant to verify that the benchmark can run on your system
* `dotnet fsi main.fsx upper` - to run a variant with largest product count
## Filling database / running endpoints
It is possible also to run also parts of the scripts, using the F# interactive console.
First, start the interactive console:
```
dotnet fsi
```
To start using the commands, you need to load the scripts
```
#load "benchmark.fsx";;
```
The next commands depends on an available folder structure, the following command creates or empties needed folders.
```
Benchmark.createAndClearWorkingDirectories ();;
```
After that, generate the dataset (replace the number with desired prod count)
```
Benchmark.generateData 20;;
```
With the generated dataset, you can easily start the database in a docker network
```
open Database
open Docker
createNetwork benchmarkNetwork
startDatabaseContainer Databases.MsSql
;;
```
Now, with everything available, it is possible to start the endpoint
```
open Endpoints
eviEndpoint.Start Databases.MsSql
;;
```
And to run a benchmark on it, the arguments are:
* How many runs will be executed
* Suffix for the log and output files
* How many parallel clients will be used
* Which endpoint will be benchmarked
* A flag whether the full log should be stored
```
Benchmark.runSingleBenchmark 128 "log-file-suffix" 1 eviEndpoint false;;
```
In the end, you have to also stop the docker containers and remove network:
```
stopAndRemoveContainer eviEndpoint.DockerName
stopAndRemoveContainer databaseDockerName
removeNetwork benchmarkNetwork
;;
```
To close the interactive session, just write:
```
#quit;;
```
## Running the benchmark in Azure
For simplicity, we have also included a simple way how to run the benchmark using hw placed in Azure: [guidance](https://github.com/mchaloupka/r2rml-benchmark/tree/master/azure).