We switched from memcache-repcache for tickets and /MySQL for service registry to using replicated ehcache for the ticket registry and JSON service registry (We manage with puppet, but git/cron would work just as well.) This sits behind an apache load balancer. From a clustering standpoint it works great for us.
On Mon, Oct 13, 2014 at 12:28 PM, Andrew Morgan <[email protected]> wrote: > On Mon, 13 Oct 2014, Benito J. Gonzalez wrote: > > Dear Community, >> >> We are looking to build a High Availability CAS solution for our campus >> rather quickly. We currently have two CAS app VMs sharing tickets with a >> single database as a single point of failure. We have F5 in front to >> load-balance traffic. >> >> What are some high level architectures for HA you have implemented? >> > > We have a 2-node CAS cluster behind a load balancer. The ticket registry > is stored in memcache+repcache. CAS connects to memcache on the localhost > and it is replicated to the other node. > > The service registry is stored in MySQL. CAS continues to process > authentications while MySQL is down (note: see issue CAS-1458), although > you cannot make changes to services, of course. > > Andy > > > -- > 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 > -- --------------------------------------------- Casey Feskens <[email protected]> Associate Director of Systems Services Willamette Integrated Technology Services Willamette University, Salem, OR Phone: (503) 370-6950 --------------------------------------------- -- 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
