users@glassfish.java.net

Re: GFv2: Resync of failed cluster nodes

From: <glassfish_at_javadesktop.org>
Date: Fri, 18 May 2007 08:09:47 PDT

Hi Gopal,

I retried the scenario with the latest b47, Sun Web Server 6.1SP7 + latest lb plugin drops. I see the same problematic result:
- failover to a different instance (#2) works great
- re-start killed instance (#1) and wait until its up an running
- access the app on #2, kill #2
and from time to time I can see in the log files that a new session gets created.
I also tried timeBased Replication versus webMethod but without a difference. I also tried 3 node cluster with the same result.

-Daniel
[Message sent by forum member 'barz26' (barz26)]

http://forums.java.net/jive/thread.jspa?messageID=217849