Gossip-based aggregation in large dynamic networks

@article{Jelasity2005GossipbasedAI,
  title={Gossip-based aggregation in large dynamic networks},
  author={M{\'a}rk Jelasity and Alberto Montresor and {\"O}zalp Babaoglu},
  journal={ACM Trans. Comput. Syst.},
  year={2005},
  volume={23},
  pages={219-252}
}
As computer networks increase in size, become more heterogeneous and span greater geographic distances, applications must be designed to cope with the very large scale, poor reliability, and often, with the extreme dynamism of the underlying network. Aggregation is a key functional building block for such applications: it refers to a set of functions that provide components of a distributed system access to global information including network size, average load, average uptime, location and… 
Decentralized Ranking in Large-Scale Overlay Networks
TLDR
This paper proposes T-Rank, a robust and completely decentralized algorithm for solving the ranking problem with minimal assumptions, and presents extensive empirical results that suggest near logarithmic time complexity, scalability and robustness in different failure scenarios.
A generic and adaptive aggregation service for large-scale decentralized networks
TLDR
DIAS, the Dynamic Intelligent Aggregation Service is an agent-based middleware that addresses issues with a holistic approach: an efficient availability of the distributed information in every node of the network that enables the simultaneous computation of almost any aggregation function.
Epidemic-Style Global Load Monitoring in Large-Scale Overlay Networks
TLDR
A gossip-based algorithm for monitoring large-scale distributed systems is presented and its efficiency in a simulated environment provided by OverSim is analyzed.
Fast Estimation of Aggregates in Unstructured Networks
TLDR
This paper introduces Extrema Propagation, a probabilistic technique for distributed estimation of the sum of positive real numbers, which relies on the exchange of duplicate insensitive messages and can be applied in flood and/or epidemic settings.
Flow updating: Fault-tolerant aggregation for dynamic networks
Dependability in Aggregation by Averaging
TLDR
This paper reveals some dependability issues of aggregation algorithms based on iterative averaging techniques, and proposes and evaluates two new versions of the Push-Pull Gossip protocol, which solve its message interleaving problem (evidenced even in a synchronous operation mode).
Using less links to improve fault-tolerant aggregation
TLDR
Flow Updating is an aggregation algorithm that provides an accurate estimate at all nodes, and it is by design immune to message loss, so this work proposes a simple heuristic that extends the initial version of Flow Updating by locally ignoring some communication links between neighbors while keeping it's fault-tolerance properties.
Tweaking gossip algorithms for computations in large-scale networks
  • S. Dulman, E. Pauwels
  • Computer Science
    2015 IEEE Tenth International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP)
  • 2015
TLDR
It is shown that for large-scale networks, SFC proves to be a surprisingly fast and robust aggregation method, easily extendable to allow self-stabilization.
Kernel Based Mechanisms for High Performance I/O
TLDR
This thesis introduces and formally analyze two sets of randomized gossip protocols for collaborative information spreading in P2P networks, and provides an efficient solution that exhibits the inherent robustness and scalability of gossip.
Extrema Propagation: Fast Distributed Estimation of Sums and Network Sizes
TLDR
This paper introduces Extrema Propagation, a probabilistic technique for distributed estimation of the sum of positive real numbers, which relies on the exchange of duplicate insensitive messages and can be applied in flood and/or epidemic settings.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 36 REFERENCES
Gossip-based computation of aggregate information
  • D. Kempe, A. Dobra, J. Gehrke
  • Computer Science
    44th Annual IEEE Symposium on Foundations of Computer Science, 2003. Proceedings.
  • 2003
TLDR
This paper analyzes the diffusion speed of uniform gossip in the presence of node and link failures, as well as for flooding-based mechanisms, and shows that this diffusion speed is at the heart of the approximation guarantees for all of the above problems.
The Peer Sampling Service: Experimental Evaluation of Unstructured Gossip-Based Implementations
TLDR
This paper presents a generic framework to implement reliable and efficient peer sampling services, which generalizes existing approaches and makes it easy to introduce new ones, and shows that all of them lead to differentpeer sampling services none of which is uniformly random.
Scalable fault-tolerant aggregation in large process groups
TLDR
A technique to impose an abstract hierarchy on such large groups, describing how this hierarchy can be made to mirror the network topology is proposed and presented, and a protocol based on gossiping is presented that uses this hierarchical technique.
The Importance of Aggregation
  • R. V. Renesse
  • Computer Science
    Future Directions in Distributed Computing
  • 2003
TLDR
This paper defines aggregation as the ability to summarize information as the basis for scalability for many, if not all, large networking services, and develops an aggregation facility, called Astrolabe, for use by networked services.
Detection and Removal of Malicious Peers in Gossip-Based Protocols∗
TLDR
A general framework for solving the security problem in P2P protocols based on gossip-based communication in an unstructured communication topology is introduced and some initial ideas towards solutions are suggested.
Large-Scale Newscast Computing on the Internet
TLDR
The newscast model of computation for large-scale computing on the Internet is introduced, a lazy fully distributed information propagation protocol among the participants which is responsible for membership management and communication and some interesting application areas are suggested.
A scalable distributed information management system
TLDR
This work designs, implements and evaluates a Scalable Distributed Information Management System (SDIMS) that leverages Distributed Hash Tables (DHT) to create scalable aggregation trees, achieves isolation properties at the cost of modestly increased read latency in comparison to flat DHTs, and gracefully handles failures.
Epidemic information dissemination in distributed systems
TLDR
Four key problems: membership maintenance, network awareness, buffer management,buffer management, and message filtering are described and some preliminary approaches to address them are suggested.
A Gossip-Style Failure Detection Service
TLDR
A new protocol based on gossiping is described that does scale well and provides timely detection, and is extended to discover and leverage the underlying network topology for much improved resource utilization.
A Modular Paradigm for Building Self-Organizing Peer-to-Peer Applications
TLDR
The goal of this paper is to promote a modular paradigm in which well-understood, predictable components with clean interfaces can be combined to implement arbitrarily complex functions, by describing the initial experiences in this direction.
...
1
2
3
4
...