Databases Reference
In-Depth Information
FIGURE 6.3: High-level views of a P2P computing model (left) and a client-
server model (right).
far beyond this simple function. It might be possible to use P2P for complex
operations, such as collaborative processing, in large-scale problems.
Some of the goals for the P2P system described in [78] include the following:
• Cost sharing/reduction: Reduces the operational costs associated with
large and centralized resources by distributing these resources to peers.
• Improved scalability/reliability: The scalability and reliability of P2P
systems are enhanced using innovative algorithms for resource discovery
and search.
• Resource aggregation and interoperability: Because of its ability to ag-
gregate resources, computationally expensive applications can be e
-
ciently performed in a P2P system, e.g., distributed systems such as
SETI@home.
• Increased autonomy: Each node in a P2P system has full autonomy on
the resources that reside on the network, without depending on a central
server.
• Anonymity/privacy: Users can hide their profiles.
• Dynamism: The computing environment within a P2P system is as-
sumed to be highly dynamic, i.e., resources, such as compute nodes,
continuously enter and leave the system.
• Enabling ad-hoc communication and collaboration: The dynamism of
P2P systems allows for ad hoc communication and peer-to-peer collab-
oration.
