> I'm trying to figure out a way to have a backup CAS server in case of a > failover. ... I just dont want to have a global > denial of service in case the server fails.
If you don't care about losing authenticated state, which the CAS clustering guide deals with, then you can simply set up two CAS instances with the default in-memory ticket storage and use some kind of load balancing strategy to route requests to a standby machine if the primary is unavailable. The load balancer will need some kind of health-checking capability to determine whether the primary is responding, but that is a common feature. From our experience layer 7 health checks are the best; for example, an HTTP probe on the /login URI that expects a response code of 20x. M -- You are currently subscribed to [email protected] as: [email protected] To unsubscribe, change settings or access archives, see http://www.ja-sig.org/wiki/display/JSG/cas-user
