Weblogic JSessionId uniqueness

Hi all,

I have two linked questions about the uniqueness of the JSsssionId, I'm interested especially on the sessionId part (i.e. before node identifiers and timestamp):

1) How is the unique generation of sessionIds guaranteed in a cluster of weblogic Servers ? (if it is, I imagine statistically).

2) Which algorith is used to compute it? In case of a pseudo random algorithm it would be interesting to have an idea about the probability of having clashes.

Thanks.