Redis is another in-memory hash store with disk persistence. It's different from memcached or Tokyo Cabinet because it has native list and set types. Even better, it has atomic operations like set intersection, increment/decrement, sort (!) and so on. This makes it suitable for implementation of real applications without traditional relational databases.
If you want to try it with perl, head over to Redis perl bindings and redis from git repository at github. It supports tie to hash or arrays if you don't want to use Redis API directly.
I just finished implementing current version of protocol (which will change soon it seems) and I would like to implement some sample scripts before I push it to CPAN near you.