300 likes | 452 Views
Unstructured overlays: construction, optimization, applications. Anne-Marie Kermarrec Joint work with Laurent Massoulié and Ayalvadi Ganesh. N nodes in a group. Each node gossips new messages to K other nodes chosen at random .
E N D
Unstructured overlays: construction, optimization, applications Anne-Marie Kermarrec Joint work with Laurent Massoulié and Ayalvadi Ganesh
N nodes in a group. Each node gossips new messages to K other nodes chosen at random. How large should K be so that ever node receive the message with high probability? Stronger than requiring that nearly 100% get the message with high probability Epidemic protocolsEpidemic multicast 3 4 0 1 9 8 5 2 6 7
Epidemic protocolsPerformance • Modelled as a random graph • Erdos and Renyi result applies to connectivity of undirected graph. • Sharp threshold at log N. • Main results If K= log(N) + c, the probability that every node is reached is exp(−exp(−c)). Result applies if mean out-degree is log(N) + c, irrespective of the degree distribution Use of these results to parameterize protocols
Epidemic protocolsPerformance 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Fanout 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Proportion of atomic multicast Proportion of infection in non atomic multicast
Epidemic protocolsReliability 99.98 99.94 100 90 80 70 60 50 40 30 20 10 0 0% 10% 20% 30% 40% 50% Percentage of faulty nodes Proportion of atomic multicast Proportion of infection in non atomic multicast
Epidemic protocolsResearch issues • Gossip-based algorithms • Scalable: load on each node grows logarithmically with group size • Highly Reliable: • Probabilistic guarantees • Proactive • Graceful degradation in the presence of failures • Major drawbacks • Non-scalable membership protocol • Oblivious to network topology • Generates a large number of messages in non faulty environments
Partial membership Self-organizing membership protocols Decentralized Self-set fanout Topologically awareness Reducing Traffic Epidemic protocolsAgenda SCAMP LOCALISER TREE-BASED APPLICATION-LEVEL MULTICAST
Epidemic protocolsSCAlable Membership Protocol • Partial knowledge: Each node has only a partial knowledge of the membership: local view • Adequate for reliability: O(log(n)) • Self-organizing and fully decentralized: size of local views converges to (c+1) log(N) • Membership management • Graph growth • Graph maintenance
Epidemic protocolsJoin algorithm P=1/sizeof view (1-P) Join request forwarded contact new Join request to a random member
Epidemic protocolsSubscription algorithm 76 Local view 1 4 5 6 4 0 6 7 2 36 6 6 1 2 6 0 8 3 1 7 0 1 56 6 6 5 8 7
Epidemic protocolsAverage case analysis D(n) : Average size of local view with n nodes present. Subscription adds D(n)+1 directed arcs, so (n+1) D(n+1) = n D(n) + D(n)+1 Solution of this recursion is D(N) = D(1) + 1/2 + 1/3 + …+ 1/N log(N)
Epidemic protocolsGraph maintenance: Redirection • Analysis assumes that new nodes subscribe to a random pre-existing node. • Redirection • Use of weights reflecting the connectivity of the graph • A node receiving a new subscription request may redirect it to a member of its local view . • Subscription request performs random walk on membership until it is eventually kept at some node. • Stopping rule: random walk is close to uniform on all nodes.
Epidemic protocolsPerformance • Convergence of view size • Confirms theoretical analysis • Impact of redirection • Impact of lease • Reliability • Comparison with traditional gossip • Attests to the “good” quality (uniformity) of views
Epidemic protocolsReliability 1 0.9 0.8 Full membership 0.7 SCAMP Proportion of nodes reached by the multicast 0.6 0.5 0.4 0.3 0.2 0.1 0 0% 10% 20% 30% 40% 50% 60% 70% Percentage of node failures
Degree balancing in Scamp Mean = 18
Rewiring • Balanced number of neighbours • Topology-aware • Minimize the cost function di: degree of node I (neighbours) c(i,j): cost of transmission i→j (e.g. distance) • Keeping the number • of edges fixed • Local knowledge
Select “open triangle” i—j—k at random; Evaluate locally cost of rewiring to i—k—j : Change to i—k—j with probability Distributed rewiring rule k k j j i i
Experiments • Simulations • GT Topologies • Overlay created by Scamp • Metrics • Mean distance to neighbors • Maximum and distribution of degree • Graph connectivity • Average on 100 simulations
Impact on the degree Scamp-GT Topology, 50,000 Nodes-mean degree = 18
Impact on the distance to neighbours 50,000 Nodes
Application-level multicast Good quality underlying overlay Tree-based multicast Source initiates the tree building by flooding A node takes as a parent the first node it hears from Small-world optimization Diameter (in hops) Failure resilience
Conclusion • Reshaping unstructured into loosely structured overlays: degree balancing and locality • Support for efficient application-level multicast • More work on network load/overhead • Others reshaping metrics