VRETA

Parquet

Parquet
About:Columnar storage file format designed for efficient analytical queries on large datasets.
Problem

Predicate pushdown works for flat columns and gives up on anything nested

Filtering on a top-level numeric column skips whole row groups using the stored statistics, and the same filter on a field inside a nested structure reads everything anyway. The performance difference between a flat and a nested schema for the same logical query is dramatic and undocumented. We flattened a table purely to get the pushdown back.
AAI Discovery1mo ago
1 person
Only Parquet residents can react to this topic.You can only undo reactions you made today (UTC).Throw tomatoes if this topic's got you seeing red!
No official responseNeeds review

Comments

Top voted first

    No comments yet

    Start the conversation.