SPARQL endpoint simple cache

This sets a simple cache for a SPARQL endpoint.

Caching SPARQL request can greatly improve the response of the SPARQL endpoint. Virtuoso is for instance only caching internal structures used in the calculation of the response. It does not implement a query cache.

However, SPARQL endpoint caching comes with a challenge. Since SPARQL queries can create a GET HTTP request larger than what is accepted by most (proxy) services, many of the applications use POST to get around the limitation even if the data is readonly and fully cacheable. Therefore standard url proxy caching is not effective. This proxy cache is a special instance suited for READONLY sparql endpoints.

It is build upon Nginx because Apache 2.4 does not allow to specify caching POST http requests.

Deployment structure

The service is organized as follows:

@build-time

• /config : the configuration files

@runtime

• /logs : the logs

Architectural embedding

The service connects to other services, which should be declared as accessible hosts:

• sparql-endpoint-service : SPARQL endpoint

The service URLs are given by the following environment variables

Environment variable	default value
ENV_SPARQL_ENDPOINT_SERVICE_URL	http://sparql-endpoint-service:8890/sparql

Configuration

Environment variable	default value
ENV_SERVICE_URL	sparql.data.vlaanderen.be
ENV_SUCCESS_REQUEST_CACHE_DURATION	60m
ENV_FAILED_REQUEST_CACHE_DURATION	1m

Execution

A typical start in production setting will be

docker run -d -p 80:80 
    --add-host sparql-endpoint-service:<ip/hostname>
    -e ENV_FAILED_REQUEST_CACHE_DURATION=2m
    --name=scache
    -v /persistentstorage/logs/:/nginx/logs
    bertvannuffelen/simplesparqlcache

Note the public exposure of the logs on persistent storage. It is a good practice to ensure that the logs are stored on a safe location. If the service has to change, the logs are kept for future problem resolution.