[IMPALA-8759] Use double precision for HLL

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: Impala 3.2.0
• Fix Version/s: Impala 3.4.0
• Component/s: Backend
• Labels: perf ramp-up
• Epic Color: ghx-label-3
• Assignee: Wenzhe Zhou
• Reporter: Peter Ebert
• Created: 12/Jul/19 22:08
• Updated: 26/Feb/20 23:55
• Resolved: 26/Feb/20 21:38

Description

For /be/src/exprs/aggregate-functions-ir.cc the finalize function uses a float which is only capable of 6-9 digits of precision.  More accurate estimates for larger cardinalities (beyond 999,999) should be possible with double precision.  Another c++ implementation uses double as well https://github.com/dialtr/libcount