Hi,
we are using an RDBMS 19 for a datawarehouse. We run an ETL that drops everything and build it from ground up. But the first time, the statistics job run into the ETL, so there many tables where empty at this point.
So the statistics show num_rows=0, although the table consists of 1 oder 2 million datasets. A new ETL run now takes forever, because of the failed statistics.
Does it make sense to drop the statistics, let run the ETL, and rebuild it aftwards? Thanks in advance.