At the very end of the Pastebin design walkthrough, the author briefly mentions:
We can store our contents in an Object Storage like Amazon’s S3. Whenever we feel like hitting our full capacity on content storage, we can easily increase it by adding more servers.
As I understand it, they are proposing that you add additional object storage servers to provide more space; not to provide CPU or something like that. Meaning that the data is unique on each server; not replicating.
So how does your API server know what object storage server to look for to find the paste? Are we associating a server ID with the paste in the meta database? It seems weird to put a load balancer between the API server and the object store if the objects stores are not all the exact same.