activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Brian (JIRA)" <jira+amq...@apache.org>
Subject [jira] [Comment Edited] (AMQNET-391) Add buffered stream in TcpTransport to avoid excessive packet transmision
Date Wed, 11 Jul 2012 12:26:34 GMT

    [ https://issues.apache.org/jira/browse/AMQNET-391?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13411220#comment-13411220
] 

Brian edited comment on AMQNET-391 at 7/11/12 12:25 PM:
--------------------------------------------------------

Extra packets are still being generated when stream becomes larger than the buffer of BufferedStream.
The default buffer size is 4096 bytes so sending a stream greater than 4096 bytes results
in 1 or more additional packets compared to when tight encoding is disabled.

If 2 extra packets are generated, the total bytes sent when tight encoding is enabled becomes
larger than not using tight encoding.
                
      was (Author: pebrian27):
    Extra packets are still being generated when stream becomes larger than the buffer of
BufferedStream. The default buffer size is 4096 bytes so sending a stream greater than 4096
bytes results in 1 or more additional packets compared to when tight encoding is disabled.
                  
> Add buffered stream in TcpTransport to avoid excessive packet transmision
> -------------------------------------------------------------------------
>
>                 Key: AMQNET-391
>                 URL: https://issues.apache.org/jira/browse/AMQNET-391
>             Project: ActiveMQ .Net
>          Issue Type: Improvement
>          Components: ActiveMQ
>    Affects Versions: 1.5.5
>            Reporter: Timothy Bish
>            Assignee: Timothy Bish
>            Priority: Minor
>             Fix For: 1.5.6, 1.6.0
>
>
> Currently the TcpTransport utilizes a raw NetworkStream which is okay when loosmarshalling
is enabled and size prefix is turned on but when either of those is not true the marshaling
code writes directly to the stream and can result in a lot of small Tcp packets getting transmitted
for each message instead of one larger more efficient write.  

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message