On 4/20/07, Jean-Baptiste Quenot <[EMAIL PROTECTED]> wrote:
Which one of these 4 issues is a security issue?

That one is already fixed (WICKET-476).

Basically it is Wicket not cleaning up a thread local at the end of a
request under very rare circumstances, messing up wicket's
administration and *can* result in a page being displayed coming from
the history of another session (but doesn't necessarily have to).

Whether you will be experiencing this behavior depends on a number of factors:
* you have to trigger the non-cleaning up
* thread pool size
* type of requests
* activity of users
* session timeouts
* pagemap size

This is not remote exploitable: it isn't triggered by a particular
external stimulus, and you can't rely on the behavior (the thread pool
of the application server determines this).

The thread local clearing logic has been altered so that it *always*
is cleared at the end of the request.

In our system we were able to track the problem (though not find a
particular request or usecase to be the source of the troubles). Since
we applied this patch, we haven't seen the cross session pollination
happen any more (we're at 1.3-something).

Martijn

--
Learn Wicket at ApacheCon Europe: http://apachecon.com
Join the wicket community at irc.freenode.net: ##wicket
Wicket 1.2.5 will keep your server alive. Download Wicket now!
http://wicketframework.org

Reply via email to