[
https://issues.apache.org/jira/browse/PARQUET-2251?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mars updated PARQUET-2251:
--------------------------
Description:
In parquet pageV1, even all pages of a column are encoded by dictionary, it
will still generate BloomFilter. Actually it is unnecessary to generate
BloomFilter and it cost time and occupy storage.
Parquet pageV2 doesn't generate BloomFilter if all pages of a column are
encoded by dictionary,
was:In parquet pageV1, even all pages of one column are encoded by dictionary
> Avoid generating Bloomfilter when all pages of one column are encoded by
> dictionary
> -----------------------------------------------------------------------------------
>
> Key: PARQUET-2251
> URL: https://issues.apache.org/jira/browse/PARQUET-2251
> Project: Parquet
> Issue Type: Bug
> Reporter: Mars
> Priority: Major
>
> In parquet pageV1, even all pages of a column are encoded by dictionary, it
> will still generate BloomFilter. Actually it is unnecessary to generate
> BloomFilter and it cost time and occupy storage.
> Parquet pageV2 doesn't generate BloomFilter if all pages of a column are
> encoded by dictionary,
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
