Database Reference
In-Depth Information
Data replication
Data replication in DynamoDB is handled by a special type of node called a
coordinator
node
. The replication factor (N) is configured per instance, which is nothing but the num-
ber of replications of a certain key. The coordinator node first keeps the key on the local
machine and then replicates it to N minus 1 successor machine in the clockwise direction,
as shown in the following diagram:
The previous diagram says that if you have a key, say, K1, which falls in the range
Node1-Node2, then replicas of the key would be stored on two successor nodes of Node1.
The list of nodes saving a particular key is called a preference list. While creating this list,
we need to make sure that we don't have any virtual nodes, as they might be referencing the
same node.
