Kafka and MemSQL

Should the kafka partitions use the same partition algorithm as used in memsql?
I am using shard key to distribute data.
As the documentation says that each kafka partition is mapped to a leaf node and the leaf node pulls the data directly.
What happens if the algorithm is not same?
I do not want to use empty shard key ().
Could anyone please guide me here!?

MemSQL intelligently maps Kafka partitions to MemSQL partitions and you don’t have to worry about matching partition counts, shard keys, or anything else. It should just work