The filter's underlying SQL query was using a DISTINCT to eliminate duplicate cluster checkins before running aggregate functions (like MIN and MAX) on all the checkins. That was wrong for the below reasons:
it was incorrect to use the DISTINCT keyword for a few reasons:
Postgres required that the Primary Key of the clusters table was also in the GROUP BY statement
The DISTINCT was eliminating rows that needed consideration by the aggregate functions below
The filter's underlying SQL query was using a
DISTINCT
to eliminate duplicate cluster checkins before running aggregate functions (likeMIN
andMAX
) on all the checkins. That was wrong for the below reasons:it was incorrect to use the
DISTINCT
keyword for a few reasons:GROUP BY
statementDISTINCT
was eliminating rows that needed consideration by the aggregate functions belowFixes #104