[
https://issues.apache.org/jira/browse/IMPALA-11460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17760834#comment-17760834
]
Daniel Becker commented on IMPALA-11460:
----------------------------------------
I ran a benchmark where a global threadpool of 4 threads and queue length of 8
was used for async codegen (see
[CallableThreadPool|https://github.com/d-becker/impala/blob/876829951dc23ac044e79c3ed3088207d812d3bb/be/src/util/thread-pool.h#L221]).
Added the results in {{{}global_threadpool4-8_async_benchmark_results.txt{}}}.
It shows a slightly bigger improvement (-30.34%). TPCH-Q1 still shows
regression but it is smaller, +4.47%. In contrast to the previous benchmark
(without a threadpool), significant variability was detected for TPCH-Q2 andÂ
TPCH-Q11.
> Enable ASYNC CODEGEN by default
> -------------------------------
>
> Key: IMPALA-11460
> URL: https://issues.apache.org/jira/browse/IMPALA-11460
> Project: IMPALA
> Issue Type: Improvement
> Reporter: Abhishek Rawat
> Assignee: Daniel Becker
> Priority: Major
> Attachments: Async_benchmark.txt,
> global_threadpool4-8_async_benchmark_results.txt
>
>
> Would be good to do some additional testing and address any gaps and enable
> the feature by default.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
