Does using Memcache lead to a change in database query design? For instance, in my application each user has an ignore list, where they can specify users they'd like to block. Once blocked neither person can see the existence of the other in the system anymore. Now suppose user A (who has blocked user B) is listing user C's friends, which contains user B. My SQL query would be something like select from Friends where ID=C and where none of these members are on A's ignore list.
Now, if I want to start storing each user's ignore list in Memcache, it would mean I just grab C's friends list, then filter out the ignored users in memcache using program logic. Is this a good way of using Memcache? Or is there a better way?
