Mailing-List: contact dev-help@directory.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Apache Directory Developers List" <dev@directory.apache.org>
Message-ID: <32189614.1186001873070.JavaMail.jira@brutus>
Date: Wed, 1 Aug 2007 13:57:53 -0700 (PDT)
From: "Martin Alderson (JIRA)" <jira@apache.org>
To: dev@directory.apache.org
Subject: [jira] Resolved: (DIRSERVER-998) Messages time out if acknowledged
 too quickly
In-Reply-To: <32224433.1184528284682.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


     [ https://issues.apache.org/jira/browse/DIRSERVER-998?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Martin Alderson resolved DIRSERVER-998.
---------------------------------------

    Resolution: Fixed
      Assignee: Martin Alderson

Fixed by scheduling the messages for timeout before they are sent.

> Messages time out if acknowledged too quickly
> ---------------------------------------------
>
>                 Key: DIRSERVER-998
>                 URL: https://issues.apache.org/jira/browse/DIRSERVER-998
>             Project: Directory ApacheDS
>          Issue Type: Bug
>          Components: mitosis
>    Affects Versions: 1.5.0
>            Reporter: Martin Alderson
>            Assignee: Martin Alderson
>            Priority: Minor
>             Fix For: 1.5.1
>
>
> Every so often a message (a login or logEntry message) will be written to a peer (by a 'client') and an acknowledgement is sent back before the sender knows that the original message has been sent.  We do not allow any more messages to be sent until the last sent message has been acknowledged or has timed out (which takes 60 seconds by default).
> In the code, the points of interest are ReplicationClientContextHandler#messageSent, ReplicationClientContextHandler#messageReceived and ReplicationClientContextHandler#beginReplication (which only begins replication if there are no unacknowledged messages).
> This isn't a major problem as replication will continue after the message has timed out, but it slows things down and is especiually bad for tests which expect replication to succeed within a few seconds.
> When this issue is fixed ReplicationServiceITest#testTwoWayBind should be re-enabled.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.