Hi Luke, This is the current expected behavior, as the Cache is itself responsible for dictating whether or not local or remote memory should be used, since that is typically in the realm of a caching products/APIs (i.e. most Caching projects/products handle primary/secondary caching strategies better than Shiro could).
That being said, I'm completely open to supporting a threadlocal-based implementation, and I do think it would make sense. I'd be happy to collaborate on this with you if you're interested. Please open a Jira issue if you'd like to keep track of it! Best, -- Les Hazlewood CTO, Katasoft | http://www.katasoft.com | 888.391.5282 twitter: http://twitter.com/lhazlewood katasoft blog: http://www.katasoft.com/blogs/lhazlewood personal blog: http://leshazlewood.com
