[
https://issues.apache.org/jira/browse/FLINK-19595?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17224522#comment-17224522
]
liuxiaolong commented on FLINK-19595:
-------------------------------------
https://github.com/Coderlxl/flink/commit/b211f4830a7301bf9283a6d37209000b176913ad
> Flink SQL support S3 select
> ---------------------------
>
> Key: FLINK-19595
> URL: https://issues.apache.org/jira/browse/FLINK-19595
> Project: Flink
> Issue Type: Improvement
> Components: FileSystems, Table SQL / Ecosystem
> Reporter: liuxiaolong
> Priority: Major
>
> Flink is based on S3AInputStream.java to select datas stored in Tencent COS,
> it will call the getObject function of AmazonS3Client.java.
> Now, Tencent COS have already support to pushdown the CSV and Parquert file
> format.
> In these cases, using getObject to select datas will wastes a lots of
> bandwidth.
> So, I think Flink SQL should support S3 Select, to reduce the waste of
> bandwidth.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
