No, this is not an HA-JNDI thing.
(B
(BThe exception you mention in your first e-mail is not complete enough: I
(Bonly see your stacktrace, not what happens in JBoss.
(B
(BAs for restarting JBoss, did you restart all nodes of the cluster or only
(Bone? you had to restart the clients as well? the current error simply means
(Bthat the specific SFSB instance is no more stored and that you must
(Bre-create one i.e. Remote newremote = sfsbHome.create();
(B
(BCheers,
(B
(B
(B
(B Sacha
(B
(B> -----Message d'origine-----
(B> De?: [EMAIL PROTECTED]
(B> [mailto:[EMAIL PROTECTED]]De la part de Nick Mills
(B> Envoye?: lundi, 27 janvier 2003 11:54
(B> A?: [EMAIL PROTECTED]
(B> Objet?: [JBoss-user] RE: [JBoss-user] SFSB:$B!!(Bfailed to recover
(Bsession
(B>
(B>
(B> Hello Sacha,
(B> My client (tomcat and a console test client) were unable to call methods
(B> on the SFSB without the error being generated:
(B> java.rmi.NoSuchObjectException: Could not activate; failed to recover
(B> session (session has been probably removed by session-timeout)...
(B>
(B> Restarting jboss was how I fixed this and allowed my clients to call ejb
(B> methods. It seems like at the timeout "something bad happened" and jboss
(B> required a reboot. I was using mostly standard cluster settings - except
(B> I did change the auto-discovery port in
(B> org/jboss/ha/jndi/HANamingService.java. Could this have caused the
(B> problem? If so I will test it with the standard port.
(B>
(B> Regards,
(B> nick
(B>
(B>
(B>
(B> -----Original Message-----
(B> From: [EMAIL PROTECTED]
(B> [mailto:[EMAIL PROTECTED]] On Behalf Of Sacha
(B> Labourey
(B> Sent: Monday, January 27, 2003 5:54 PM
(B> To: [EMAIL PROTECTED]
(B> Subject: [JBoss-user] RE: [JBoss-user] RE: [JBoss-user]
(BSFSB:$B!!(Bfailed to
(B> recover session
(B>
(B> Hello Nick,
(B>
(B> > In response have a couple more questions. From what you said it
(B> appears
(B> > that I need to restart my server every time the HASession times out -
(B>
(B> ?!? What do you mean by "I need to restart my server" ?!? What will
(B> timeout
(B> are the server sessions that haven't been activated (invoked) in the
(B> last 30
(B> minutes, that's all
(B>
(B> > which on the default settings is if the server is not used for more
(B> than
(B> > 30 minutes?
(B> >
(B> > If so can I safely change the default BeanCleaningDelay to say 10
(B> hours?
(B>
(B> In this case, the check will only be performed every 10 hours.
(B>
(B> > I am seriously considering your advice to upgrade to 3.0.6.
(B> Reliability
(B> > is key for us. Is 3.0.6 clustering production ready? Where can I get
(B> the
(B> > change logs? I am interested to know the 3.0.6/3.0.5 changes.
(B>
(B> People are running in production with 3.0.6 (and did it with previous
(B> releases as well) but 3.0.6 includes many bug fixes. See the clustering
(B> forum for more information.
(B>
(B> Cheers,
(B>
(B>
(B>
(B> Sacha
(B>
(B>
(B>
(B> -------------------------------------------------------
(B> This SF.NET email is sponsored by:
(B> SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
(B> http://www.vasoftware.com
(B> _______________________________________________
(B> JBoss-user mailing list
(B> [EMAIL PROTECTED]
(B> https://lists.sourceforge.net/lists/listinfo/jboss-user
(B>
(B>
(B>
(B> -------------------------------------------------------
(B> This SF.NET email is sponsored by:
(B> SourceForge Enterprise Edition + IBM + LinuxWorld =omething 2 See!
(B> http://www.vasoftware.com
(B> _______________________________________________
(B> JBoss-user mailing list
(B> [EMAIL PROTECTED]
(B> https://lists.sourceforge.net/lists/listinfo/jboss-user
(B>
(B
(B
(B
(B-------------------------------------------------------
(BThis SF.NET email is sponsored by:
(BSourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
(Bhttp://www.vasoftware.com
(B_______________________________________________
(BJBoss-user mailing list
([EMAIL PROTECTED]
(Bhttps://lists.sourceforge.net/lists/listinfo/jboss-user
