[jboss-user] [JBoss Remoting] New message: "Re: Client hangs when get JMS connection factory"

[mingjun jiang]

User development,

A new message was posted in the thread "Client hangs when get JMS connection factory":

http://community.jboss.org/message/520969#520969

Author  : mingjun jiang
Profile : http://community.jboss.org/people/mjjiangbhr

Message:
--------------------------------------------------------------
> mailto:ron.sigal at jboss.com 编写:
>  
> Hi guys,
>  
> I've done some experimentation and it looks like we're both right.  I was right that ConnectionValidator correctly detects and reports a connection failure when the ethernet cable is pulled.  That should be true for any recent version of Remoting, either 2.2.x or 2.5.x.  When ConnectionValidator tells JBossMessaging about the failure, JBM should close the connection, which should lead to a call to MicroSocketClientInvoker.handleDisconnect() in Remoting.  Now, MicroSocketClientInvoker.handleDisconnect() closes all inactive sockets, but it doesn't close sockets that are currently involved in an invocation.  That's why the number of sockets doesn't go to zero when "timeout" is set to 0.  But if "timeout" is set to a non-zero value, these active sockets will time out and get closed.
>  
> By the way, there has been some discussion about changing the default "timeout" configuration in remoting-bisocket-service.xml to a non-zero value.
>  
> -Ron
Ron, yes, in the original post, my colleague Li Lin had already said a non-zero timeout value will close the failure sockets. If you keep the timeout value as 0 (default value), the number of sockets won't go to zero. But if we change the timeout to a non-zero value, it will cause other issue just what I said previously

--------------------------------------------------------------

To reply to this message visit the message page: http://community.jboss.org/message/520969#520969