I'm trying to run a SPARQL query on Wikidata, but it times out. I'd like to download a dump and index it in some database, so I can run local SPARQL queries using HTTP requests. I also need to support Wikidata-specific extensions like SERVICE wikibase:label. I've downloaded an RDF dump. What are the next steps?
Wikimedia has documentation on how to run your own SPARQL endpoint from one of their dumps. They also have an updater that streams updates from their servers, to keep your endpoint up-to-date.
You won't need to do anything special to support their extensions, it is included by default.
I recommend using a reverse proxy (like nginx or apache) with http auth in production, the admin dashboard is accessible by default.