Someone ran this query against one of our servers recently:
SELECT DISTINCT ?class ?property WHERE { { ?s ?property ?o }. OPTIONAL { ?s
rdf:type ?class } } LIMIT 50 OFFSET 0
(then ran it again for good measure when the first one didn't finish).
The server is running Fuseki (0.2.2) with a timeout set: ja:cxtValue
"11000,20000". My question is: does anyone know why the Fuseki timeout didn't
kill this query? It was still running 10 mins later and causing slow responses
on the server.
The data that it was running against would tend to make this query take a long
time - as the first data in the indexes is 20 million or so triples all of the
same class, so it would find its first result quickly, but would have a lot of
work to do to get to 50 different classes.
(In any case I will upgrade that server to Fuseki 0.2.4 - this is a good
reminder that I should do that).
Thanks
Bill
