Distributed hash table

Structure

Keyspace: Something like 160-bit of string.
Keyspace partitioning: Decide which obj is assigned to which node within a set of nodes (Consistent hashing or Rendezvous hashing)
Overlay network: Connects the nodes, allowing them to find the owner of any given key in the keyspace

Assume we want to index a file with given filename and data in DHT.

SHA-1 hash value of the filename is generated, which produces 160-bit string as the key k.
Send the message put(k, data) to any node participating DHT.
The message is forwarded from node to node via the overlay network until it reaches the single node responsible for the key k.
That node stores the key k and data.
On read, any client does SHA-1 hash of the filename to get the key k.
Send the message get(k) to any node participating DHT.
The message is forwarded from node to node via the overlay network until it reaches the single node responsible for the key k.
The data is retrieved.

Using f(k, nodeID) to calculate which node is closest.

Using f(k, nodeID) to calculate the weight against all servers, then pick the one has highest score(height).