Since I only recently learned about the concept of “killed index tuples”, I thought there might be some others who are not yet familiar with this interesting PostgreSQL concept. This may give you an explanation the next time you encounter wildly varying execution times for the same execution plan of the same PostgreSQL query. […]