Ahh - I see what you're saying now.

We don't use the Realm's internal cache because we already cache the
source/target objects that the Realm would use during its logic.  That
is, when a Realm does a lookup from, say, a UserDAO (as in the
previous example) - which has tenant-specific logic - the returned
User would already be cached.  No need to have the Realm cache it
again.

Cheers,

Les

On Fri, Jul 8, 2011 at 6:40 PM, Jared Bunting
<[email protected]> wrote:
> No problem.  It's a minor point really, but I think that what you're saying 
> is that the dao is essentially not deterministic.  So, if userAA from tenant 
> 1 logs in, his info is cached under "userA".  Then if userA from tenant 2 
> logs in, we register a cache hit and he gets permission info for userA from 
> tenant 1.  I may have just missed it, but I was wondering how you handled 
> that situation.  Or do you just not use the caching?

Reply via email to