The problem at hand is that users visiting the www.site.com infrastructure from behind large web proxy clusters are susceptible to random authentication failures. This is because at any time during a transaction, they may be bounced to a new proxy server, which in turn has to re-query DNS. Since the Distributed Director provides only round robin DNS and does not understand the concept of "sticky" sessions, there is a 50/50 chance that they will be sent to a different datacenter, thus losing all of their session information. The result is an authentication failure error.
Are there any solutions for this problem?
