[
https://issues.apache.org/jira/browse/ARROW-11030?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17255711#comment-17255711
]
Andy Grove commented on ARROW-11030:
------------------------------------
batch size vs query time:
16384 = 191 seconds
32768 = 63 seconds
65536 = 29 seconds (roughly the same speed as Apache Spark locally for me)
131072 = 20 seconds
> [Rust] [DataFusion] TPC-H q12 does not finish with SF=100 Parquet input
> -----------------------------------------------------------------------
>
> Key: ARROW-11030
> URL: https://issues.apache.org/jira/browse/ARROW-11030
> Project: Apache Arrow
> Issue Type: Bug
> Components: Rust - DataFusion
> Reporter: Andy Grove
> Assignee: Andy Grove
> Priority: Major
> Fix For: 3.0.0
>
>
> TPC-H q12 does not finish with SF=100 Parquet input.
> Either I have something wrong with my local setup or there is a major
> performance regression. I will investigate more after the holidays.
> It is also possible that I am getting confused about which queries and scale
> factors were working well for me before. We really need to start tracking
> benchmark results somewhere.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
