In financial businesses, all data represents some kind of time series, and the time and date columns grow constantly. All my tables partitioned by date, and each day, I add a new partition.
Analyze calculates the min/max for each column, but optimizer sometimes detects a value outside of today's range, and will sometimes misreport that only a few rows will be fetched, rather than the tens of billions that I would expect. Optimizer wants to broadcast a huge table. What's the best way to resolve this?
Optimizing Query Performance and Resource Pool Tuning
@rbankula @bat
