What are some tools you have used before to successfully profile and optimize SPARQL queries? We’re using Neptune, and hoping to measure where our bottlenecks are.
cc nabeel, Ora L.
Avoiding the OPTIONAL keyword is a big help for optimizing (see https://www.bobdc.com/blog/dropping-optional-blocks-from/). Otherwise, profiling and optimization tricks are usually specific to the SPARQL engine that you are using. It's one of the ways they distinguish themselves from each other. So, contacting Neptune folk or searching for online material on Neptune optimization is probably your best bet.
I expect vendor would have some query plan feature. Stardog and AllegroGraph can show that which is sometimes helpful in determining the slow parts of a query
The “explain” feature in Neptune is basically the first step. It can sometimes be a bit hard to read its output. Let us know if we (the Neptune team) can help.
