Edit: The service is back now. I'm working on a better caching solution to lower the number of hits on the API.
Edit #2: Okay, I have a more permanent solution. Bookflavor now serializes and caches API responses locally, then hits the cache and unserializes them if possible rather than hitting the Amazon API. Hopefully it won't exceed the limits this hour. Another advantage is that hitting the local disk cache is much faster than communicating with Amazon's servers.
Thanks for helping me load test Bookflavor everyone!