One possible cause of accumulating __shibstate_xxx cookies

[Peter Schober]

* Jon Warbrick <jw35 at cam.ac.uk> [2012-04-20 13:28]:
> We've recently managed to understand one circumstance in which 
> __shibstate_xxx cookies can end up accumulating to the point where the 
> Cookie: header gets long enough that web servers start rejecting it. I'm 
> documenting it here in case the information helps anyone else.

We simple reverted/upgraded to the "new" default config (using the
storage service) and that got rid of this ever-growing cookie.
-peter