Once IMPALA-5842 has been resolved, we should skip pages based on the page index in Parquet files.
Be smarter about I/O patterns for Parquet scan ranges
Avoid Parquet pages with too many rows + try to make them aligned
Extend test_scanners_fuzz.py with selective queries
Runtime filter : Extend runtime filter to support Min/Max values for HDFS scans
Parquet Row Group Size optimization
Write page index in Parquet files