A general query
When we run a SPARQL query for a virtual graph. Does it internally just convert it to SQL queries and run it across the connected DBs.
In which case for if we have highly structured data without too many inter relationships would you recommend Virtual graphs over an ETL data ware housing solution?
Hi Nisha,
When thinking about deciding between virtualizing or materializing data I always recommend first thinking about the frequency in which your data updates and your user's performance needs.
If your data changes frequently, a virtual graph is likely the best solution so you do not need to constantly run ETL processes to update your data.
If you need the fastest possible performance, materializing data to Stardog will always be faster than having Stardog federate queries to your virtualized sources.
A more complete set of questions to ask about the trade-offs can be found in this blog: 7 Questions to Decide if you Should Virtualize or Materialize your Data | Stardog
Thanks,
Laura