1. Super-peer Network

2. Motivation: Search in P2P Centralised (Napster) Flooding (Gnutella)  Essentially a breadth-first search using TTLs Distributed Hash Tables

3. But….all peers are not equal 20% peers have a latency of at most 70ms and 20% have a latency of at least 280ms

4. Basic Idea Real-World P2P systems for the open Internet are heterogeneous  Peer resources (Bandwidth, CPU, Memory)  Peer session-time Use Peers with better “characteristics” to provide services to other peers in the system

5. Super-Peer Definition A super-peer is a node in a peer-to-peer network that operates both as a server to a set of clients, and as an equal in a network of super-peers. [Yang/Molina '03] Super-peers have high utility relative to non super-peers, where higher utility peers are “better” at providing superpeer service(s).

6. Super-Peer Networks Exploit heterogeneity in P2P Networks by using higher utility peers to provide services A super-peer network operates exactly like a pure P2P network, and each super-peer is connected to a set of clients  Clients are connected to a single super-peer only

7. Super-Peer Architecture

8. Super peer network Cluster  A super-peer and its clients Cluster size  Number of nodes in the cluster, including the super-peer itself A pure p2p network is actually a “degenerate” super-peer network where cluster size is 1  Every node is a super-peer with no clients

9. Super peer network Super-peer keeps an index over its clients’ data  Joints Client sends metadata over its collection to its super- peer  Updates Query processing  Client -> super-peer  Super-peer -> neighbors (super-peer)

10. Super-Peer Design Issues What services do super-peers provide? Ordinary peer to super-peer connections  Redundancy, Performance, Fairness Super-Peer Connection Topology Super-peer promotion  Which/how many peers should be super-peers?

11. What Services do Super-Peer Systems provide? File Naming/Retrieval  Kazaa  E-Donkey Voice Over IP (VoIP)  Skype Anything you want!

12. Ordinary Peer to Super-Peer Connections A super-peer become a single point of failure for its cluster and a potential bottleneck To provide reliability to the cluster  K-redundant Clients send queries in a round-robin fashion

13. Ordinary Peer to Super-Peer (SP) Connections A k-redundant super-peer has much great availability than a single super-peer, however  The aggregate cost of a client joint action is k times greater than before  The number of open connections amongst super-peers increases by a factor of k 2 ! Super-peer redundancy has the surprising effect of reducing load on each super-peer, in addition to providing greater reliability

14. Super-Peer Connection Topology The type of intra-SP topology affects the type of distributed services they provide Random Network  Flooding, Random Walk DHT (don't know of any systems that do this) Gradient Topology  Gradient search

15. Search in Random and Gradient SP Topologies

16. Super-Peer Promotion Options:  1. Promote peers that meet local utility requirements  2. Promote the top 'X' percent of peers with highest utility (requires global knowledge of peer utility distribution)

17. Super-Peer Promotion using Local Knowledge Promotion algorithm makes a local decision about promotion using:  Measurements of the peer’s local utility levels for Bandwidth, CPU, Memory  A Model of Peer Utility Distributions

18. Super-Peer Election using Global Knowledge Too expensive solutions  Centralized Servers  Classic Election Algorithms Gossiping / Aggregation  Used by Gradient Topology

19. Conclusion A combination of the centralized client- server model (efficiency) and distributed search (load balancing and robustness)