Paper #: [5.1 P2P] 2
Title: Chord: A scalable peer-to-peer lookup service for Internet
applications by Stoica, etc.


PROBLEM

Previous peer-to-peer implementations have suffered from one or more of the
following problems:
*Lack of Scalability because all nodes have to know where all the other
nodes are located.
*Updates to network, such as nodes joining or leaving the network, requires
a significant amount of work in order to keep the system stable.
*Some nodes in the network have a larger resource drain than others.
*A single point of failure is possible.  For example, if one node leaves,
then the information available in that network is lost forever, unless
replication was used.
*There is no guarantee of finding data even if it is on the network because
of TTL.

NEW IDEAS AND STRENGTHS

*In order to increase scalability, Chord has each node store routing
information about a small number of other nodes in order to ensure that the
entire network is connected while at the same time minimizing the amount of
information that must be stored in each node.  This also ensures that if the
data is on the network, it will be found.
*Chord uses replication in order to ensure that data is still accessible
even if a node leaves the network without warning.  Therefore, this helps to
eliminate a single point of failure.
*When a node leaves the network its' keys are reassigned which balances the
load on the network.


WEAKNESSES AND EXTENSIONS

*Unlike some implementations, Chord does not take network distance into
consideration when determining routing.  Consequently if both of nodes that
want to connect are in the same area, such as on Georgia Tech's campus,
their queries may still be routed through Russia, which would increase the
amount of time of the query.
*The structured approach of Chord costs nodes anonymity, which is one of the
strong points of peer-to-peer networks.