Carl,

> I guess a "rogue service" that was registered with the server that could 
> somehow steal STs from a neighbor service could gain access to attributes 
> that only the neighbor could see.

Services impersonating users at other services with a replayed ticket/assertion 
was precisely the reason for this in SAML.  It was a big consideration for 
inter-organizational use.

http://docs.oasis-open.org/security/saml/v2.0/saml-sec-consider-2.0-os.pdf

In general, I prefer the “don’t make them ask for permission; make them beg for 
forgiveness” model, but there is data that even remediation and lawsuits 
couldn’t claw back.

All of this is not really a consideration as long as interoperation isn’t a 
consideration, and it's a much smaller concern for the typical CAS use cases.  
This came up most recently when thinking through how stateless you *could* make 
a CAS server.

I could stuff whatever I want into a ticket as long as I understand it when 
it’s played back to me because it’s just an opaque string to the rest of the 
universe.  The length limitations will bite pretty quickly, but all this 
*could* get serialized and farmed out.  There’s just no standard way to do so.

It’s not a security concern at that point: it’s a data replication problem.  
It’s just a question of where the data used for those checks lives.  But that’s 
even harder than security.

Thanks,
Nate.

-- 
You received this message because you are subscribed to the Google Groups "CAS 
Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
Visit this group at https://groups.google.com/a/apereo.org/group/cas-user/.

Reply via email to