How is DHT different than regular hash tables in context to data/node lookup?
This (introduction, 3rd paragraph) paper says:
First, in addition to the insertion and deletion of items, DHTs must support the insertion and deletion of buckets: as machines join and leave the network, items must be migrated to other machines and the hash function revised to reflect their new location.
This problem is also in Hashtables (for example, if you implement a hashtable in C using LinkedList and Array, then if the size of the array increased (new node added) then the data needs to be redistributed to achieve good load factor)
Second, some kind of routing protocol is usually necessary: since it is not feasible in a P2P system for every node to maintain up-to-date knowledge of all other nodes in the system, an item is looked up (or inserted) by following a sequence of routing hops through the peer-to-peer network.
Is this point talking about routing mechanism, if yes, then is this done is normal hashtable?