CAS Clustering Clarifications

[Scott Battaglia]

Terry,

If you can front CAS with a Load Balancer using Sticky Sessions you only
need to have a clusterable TicketRegistry (i.e. JBossTicketRegistry).  You
will need to reconfigure the default UniqueTicketIdGenerator to use a unique
suffix for each machine (to avoid potential ticket id collisions).

CAS 3.1 will include the JBossTicketRegistry as well as a
BerkeleyDbTicketRegistry (for longer-term tickets).  CAS 3.1 is schedule for
release at the end of June.

Thanks
-Scott
-- 
-Scott Battaglia
LinkedIn: http://www.linkedin.com/in/scottbattaglia

On 5/15/07, Terry Jeske <terry.jeske at gmail.com> wrote:
>
> Hello,
>
> I've read all the threads, the wiki and the short blurb on the site about
> clustering, but I am still a bit confused.
>
> From the JaSig site on Clustering:
> "In fact, it can be as easy as replacing one or two implementations, or as
> difficult as replacing three implementations. The difficulty depends on
> whether you use Sticky Sessions with your clustering strategy (or some other
> form of session replication). "
>
> This does not line out under what circumstances you would have to
> implement, one or all three of the interfaces. For example, we have a LB
> fronting two CAS servers, and we can use sticky sessions; so which of the
> following interfaces do we have to implement?
> FlowExecutionStorage
> TicketRegistry
> DefaultUniqueTicketIdGenerator
>
> Also, I did not see it stated as such, but does 3.1 clustering support
> built in?
>
> Thanks so much for your help.
>
> _______________________________________________
> Yale CAS mailing list
> cas at tp.its.yale.edu
> http://tp.its.yale.edu/mailman/listinfo/cas
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://tp.its.yale.edu/pipermail/cas/attachments/20070523/0c0f1811/attachment.html