commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Bogdan Drozdowski (JIRA)" <j...@apache.org>
Subject [jira] Commented: (NET-258) Implement A Keepalive Mechanism
Date Fri, 11 Feb 2011 16:53:57 GMT

    [ https://issues.apache.org/jira/browse/NET-258?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12993578#comment-12993578
] 

Bogdan Drozdowski commented on NET-258:
---------------------------------------

Right. Volatile - yes, you're correct - I forgot that. About the rest: sending NOOPs does
nothing to the data transfer (theoretically even closing the control connection politely does
nothing). The data transfer is done by Input/OutputStreams, as far as I can see. So we have
one thread reading/writing and the other sending NOOPs. The keepalive thread is not started
automatically. The user of the class decides if the thread is to be run and when should it
be stopped. Worker thread is a good idea, but the problem is that we need a reply immediately,
so we can't just queue the command, because it could break the getReply methods. The intention
of the patch is that the command sender knows that he/she won't send any commands in the nearest
future, so he/she launches the thread. So if any threads the user has won't send any commands,
the user may start the thread. I'm not trying to outsmart anything and guess if we should
interfere or not. This is just a small solution for a start
 .

> Implement A Keepalive Mechanism
> -------------------------------
>
>                 Key: NET-258
>                 URL: https://issues.apache.org/jira/browse/NET-258
>             Project: Commons Net
>          Issue Type: Improvement
>    Affects Versions: 2.0
>            Reporter: Rory Winston
>            Assignee: Rory Winston
>         Attachments: ftp-keepalive.diff
>
>
> For routers/firewalls that terminate idle connections, a separate heartbeat mechanism
may need to be implemented to keep the control connection active.
> Some potential issues:
> * Synchronization between a heartbeat write and a __getReply() on an active control connection
command;
> * Should use the NOOP command as a heartbeat signal;
> * Make the timeout configurable;
> * Default SocketImpl::setKeepAlive() wont do here.

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message