[
https://issues.apache.org/jira/browse/DIRMINA-569?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jeff Mesnil updated DIRMINA-569:
--------------------------------
Attachment: DIRMINA-569.txt
This patch contains test cases to highlight this issue.
Once the connector is connected, we wait a little so that the session starts to
be idle.
After a little time, we check we've not been notified of an exception.
The same code is used with only changing the value of IdleStatus.
The tests pass for READER_IDLE & WRITER_IDLE and fails for BOTH_IDLE
> KeepAliveFilter's BOTH_IDLE is not working as expected
> ------------------------------------------------------
>
> Key: DIRMINA-569
> URL: https://issues.apache.org/jira/browse/DIRMINA-569
> Project: MINA
> Issue Type: Bug
> Components: Core
> Environment: MINA trunk (r647676)
> Reporter: Jeff Mesnil
> Attachments: DIRMINA-569.txt
>
>
> for my code, I need to send keep-alive messages when the client IoSession is
> idle both in reading and writing.
> Using IdleStatus.BOTH_IDLE does not work as I expected it:
> When the KeepAliveFilter is notified that it is BOTH_IDLE, it sends a
> keep-alive request.
> However, sending this request means that the session is no longer WRITER_IDLE.
> In turn, this means that the filter is notified that the session is
> READER_IDLE only.
> Since it is waiting a response (from the BOTH_IDLE notification) that has not
> been received, it falsely handles it as a timeout.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
