What you're describing is Consistent Hashing. It is a mean to _assign partitions to nodes_ and it is independent of how you partition your _data_ (though it's t

What you're describing is Consistent Hashing. It is a mean to _assign partitions to nodes_ and it is independent of how you partition your _data_ (though it's typically used with hash based partitioning of th edata). At any rate, the way I understood it, the article described partitioning of the data by `key mod k`, k being the current number of nodes.