users@glassfish.java.net

Re: Glassfish & cluster, When stress test, the sessionid changed, is it a b

From: <glassfish_at_javadesktop.org>
Date: Wed, 07 Nov 2007 03:28:09 PST

When I changed the "Persistence Frequency" from "web-method" to "time-based", the chance of changing become very slow or disaplear.

So I think that the problem is because of efficiency for http session store. If save same http session frequently, there may be "synchronized" problem, the session may be set Expired. The scene may replicate sometimes.
[Message sent by forum member 'hanjingdi' (hanjingdi)]

http://forums.java.net/jive/thread.jspa?messageID=244258