[IMPALA-12631] Improve count star performance for parquet scans

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: Impala 4.4.0
• Component/s: Backend
• Epic Color: ghx-label-3
• Assignee: YifanZhang
• Reporter: YifanZhang
• Created: 14/Dec/23 09:47
• Updated: 28/Feb/24 18:29
• Resolved: 08/Feb/24 15:35

Description

The code in the backend function HdfsParquetScanner::GetNextInternal() is not efficient now. We use row group statistics instead of file meta statistics, which leads to unnecessary materialization overhead.