[
https://issues.apache.org/jira/browse/ARROW-3781?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16685945#comment-16685945
]
Wes McKinney commented on ARROW-3781:
-------------------------------------
It would definitely require some design work. In
https://github.com/apache/arrow/blob/master/cpp/src/arrow/io/buffered.cc#L104,
you would need to use a buffer pool of some kind so that if Flush is holding a
temporary buffer, that Write can write to a new buffer. In any case, it's out
of scope for this issue. Once we have file system implementations for one or
more cloud services we can use benchmarks to drive the development. In the
meantime, a mock remote file system with configurable write latency could help
with throughput tests
> [C++] Configure buffer size in arrow::io::BufferedOutputStream
> --------------------------------------------------------------
>
> Key: ARROW-3781
> URL: https://issues.apache.org/jira/browse/ARROW-3781
> Project: Apache Arrow
> Issue Type: Improvement
> Components: C++
> Reporter: Wes McKinney
> Priority: Major
> Fix For: 0.12.0
>
>
> This is hard-coded to 4096 right now. For higher latency file systems it may
> be desirable to use a larger buffer. See also ARROW-3777 about performance
> testing for high latency files
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
