Improve performance when filters in RowGroupFilter can match exactly
- Lingua principale
- Java
- Stelle
- 3.1k
- Fork
- 1.6k
- Merge medio
- 3g 12h
- PR unite (30g)
- 33
Descrizione
If we can accurately judge by the minMax status, we don’t need to load the dictionary from filesystem and compare one by one anymore.
Similarly , Bloomfilter needs to load from filesystem, it may costs time and memory. If we can exactly determine the existence/nonexistence of the value from minMax or dictionary filters , then we can avoid using Bloomfilter to Improve performance.
For example,
1. read data greater than `x1` in the block, if minMax in status is all greater than `{}x1{`}, then we don't need to read dictionary and compare one by one.
1. If we already have page dictionaries and have compared one by one, we don't need to read BloomFilter and compare.
**Reporter**: [Mars](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=miracle) / @yabola
**Assignee**: [Mars](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=miracle) / @yabola
**Note**: *This issue was originally created as [PARQUET-2237](https://issues.apache.org/jira/browse/PARQUET-2237). Please see the [migration documentation](https://issues.apache.org/jira/browse/PARQUET-2502) for further details.*
Guida per i contributori
Nessuna guida per i contributori indicizzata per questo repository
Direzione di ricerca
Start by locating RowGroupFilter and tracing how minMax, page dictionaries, and Bloomfilter are evaluated. Confirm how exact minMax or dictionary results can skip later filesystem reads, then verify the behavior and performance with the relevant existing checks; the issue names no files or tests, so those must be identified first.
Scritto dal modello di indicizzazione a partire dal testo della issue.
Valutazione
- Stack tecnologico
- java
- Ambito
- data-engineering
- Tipo di issue
- Funzionalità
- Difficoltà
- 4/5
- Tempo stimato
- 3-5 giorni
- Stato di attività
- Ferma
- Chiarezza
- Abbastanza chiara
- Idoneità per principianti
- 35/100