You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hc.apache.org by "ASF subversion and git services (Jira)" <ji...@apache.org> on 2020/09/12 12:40:00 UTC

[jira] [Commented] (HTTPCORE-645) Increase default chunked request body chunk size

    [ https://issues.apache.org/jira/browse/HTTPCORE-645?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17194728#comment-17194728 ] 

ASF subversion and git services commented on HTTPCORE-645:
----------------------------------------------------------

Commit fa128ec753ab0476de9639d6c1962eb54b0462d9 in httpcomponents-core's branch refs/heads/master from Carter Kozak
[ https://gitbox.apache.org/repos/asf?p=httpcomponents-core.git;h=fa128ec ]

HTTPCORE-645: Increase blocking default chunk size from 2 KiB to 8 KiB

This value matches the default session buffer size and avoids
excessive memory copying between the chunk buffer and session
buffer.


> Increase default chunked request body chunk size
> ------------------------------------------------
>
>                 Key: HTTPCORE-645
>                 URL: https://issues.apache.org/jira/browse/HTTPCORE-645
>             Project: HttpComponents HttpCore
>          Issue Type: Improvement
>          Components: HttpCore
>    Affects Versions: 5.0.1
>            Reporter: Carter Kozak
>            Priority: Minor
>             Fix For: 5.1
>
>          Time Spent: 40m
>  Remaining Estimate: 0h
>
> Currently the default value is 2 KiB, which is relatively small as discussed in [https://github.com/apache/httpcomponents-core/pull/206#discussion_r462154831]
> Creating this ticket to discuss using a larger default value. The drawback to increasing the chunk size is that the chunk cache buffer is created for each request entity stream unlike the session buffer which is associated with a potentially persistent connection. This will result in higher allocation pressure for chunked request heavy uses, especially when the chunked request body may be much smaller than the chunk hint size. In order to work around this, the connection may create a reusable buffer based on the chunk hint which is reused across requests.
> Open questions:
> Should we increase the default chunk size?
> What value should we use?



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@hc.apache.org
For additional commands, e-mail: dev-help@hc.apache.org