Run queries from TPC-H in MapD-JDBC to measure elapsed time.
1G and 10G data worked well on 3,5,10,12 queries.
However, for 100G only, the following error occurred:
java.sql.SQLException: Query failed : Exception: Hash join failed, reason(s): Could not build a 1-to-1 correspondence for columns involved in equijoin | SlabTooBig | No equijoin expression found | No equijoin expression found
The error line was the code that executed the query.
How can I prevent Hash Join error on large amounts of data?
We’ll also measure 1TB of data in the future.