users@glassfish.java.net

Re: cluster replication problem - why ? how to debug ?

From: Shreedhar Ganapathy <Shreedhar.Ganapathy_at_Sun.COM>
Date: Thu, 21 Feb 2008 08:07:05 -0800

Hi Olivier
I will answer the Shoal part of your questions:
glassfish_at_javadesktop.org wrote:
> Yes - this is very strange.
> We will try again and rebuild all from scratch to check things.
>
> Is it possible to set debug on the javax.ee.enterprise.system.tools.synchronization package ?
> Were can we have some details on the shoal implementation in Glassfish ?
>
Details of Shoal are available at https://shoal.dev.java.net

More GlassFish specific Shoal details are available at :
http://docs.sun.com/app/docs/doc/819-3679

A bit lengthy podcast but has lot of details :
http://blogs.sun.com/glassfishpodcast/entry/episode_003_interview_with_shreedhar
http://blogs.sun.com/shreedhar/tags/shoal

> we can see some .shoal directories on the DAS cache
>
The .shoal directory structure is used by Jxta's p2p, rendezvous and
routing control. The directories are organized by server instances and
go away when each instance stops. This does not have anything to do with
Node Agent's synchronization with DAS.
Hope this contributes to isolating the problem.
> and some cached file store the timestamp of the last synchronization in milli second.
>
> regards,
> Olivier
> [Message sent by forum member 'omerlin' (omerlin)]
>
> http://forums.java.net/jive/thread.jspa?messageID=260119
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: users-help_at_glassfish.dev.java.net
>
>