Heterogeneous Gossip

Frey, Davide; Guerraoui, Rachid; Kermarrec, Anne-Marie; Koldehofe, Boris; Mogensen, Martin; Monod, Maxime; Quéma, Vivien

doi:10.1007/978-3-642-10445-9_3

Cited by 22 publications

(27 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In terms of related mechanisms, a distributed KFN graph is a form of peerto-peer network overlay. Peer-to-peer overlays have been widely applied in the past to implement distributed services, ranging from distributed storage [25,26,28], and streaming [12,22], through to pub/sub [2,24] and environmental sensing [15]. Among peer-to-peer overlays, k-nearest neighbor (KNN) overlays [17,30,4] come closest to HyFN, although they converge poorly when applied to the KFN graph construction problem, as our evaluation shows.…”

Section: Related Workmentioning

confidence: 83%

Scalable Anti-KNN: Decentralized Computation of k-Furthest-Neighbor Graphs with HyFN

Bouget

Bromberg

Taı̈ani

et al. 2017

Distributed Applications and Interoperable Systems

View full text Add to dashboard Cite

International audienceThe decentralized construction of k-Furthest-Neighbor graphs has been little studied, although such structures can play a very useful role, for instance in a number of distributed resource allocation problems. In this paper we define KFN graphs; we propose HyFN, a generic peer-to-peer KFN construction algorithm, and thoroughly evaluate its behavior on a number of logical networks of varying sizes. 1 Motivation k-Nearest-Neighbor (KNN) graphs have found usage in a number of domains, including machine learning, recommenders, and search. Some applications do not however require the k closest nodes, but the k most dissimilar nodes, what we term the k-Furthest-Neighbor (KFN) graph. Virtual Machines (VMs) placement —i.e. the (re-)assignment of workloads in virtualised IT environments— is a good example of where KFN can be applied. The problem consists in finding an assignment of VMs on physical machines (PMs) that minimises some cost function(s) [27]. The problem has been described as one of the most complex and important for the IT industry [3], with large potential savings [20]. An important challenge is that a solution does not only consist in packing VMs onto PMs — it also requires to limit the amount of interferences between VMs hosted on the same PM [31]. Whatever technique is used (e.g. clustering [21]), interference aware VM placement algorithms need to identify complementary workloads — i.e. workloads that are dissimilar enough that the interferences between them are minimised. This is why the application of KFN graphs would make a lot of sense: identifying quickly complementary workloads (using KFN) to help placement algorithms would decrease the risks of interferences. The construction of KNN graphs in decentralized systems has been widely studied in the past [17, 30, 4, 14]. However, existing approaches typically assume a form of " likely transitivity " of similarity between nodes: if A is close to B, and B to C, then A is likely to be close to C. Unfortunately this property no longer holds when constructing KFN graphs. As a result, these approaches, as demonstrated in the remainder of the paper, are not working anymore when applied to this new problem

show abstract

Section: Related Workmentioning

confidence: 83%

Scalable Anti-KNN: Decentralized Computation of k-Furthest-Neighbor Graphs with HyFN

Bouget

Bromberg

Taı̈ani

et al. 2017

Distributed Applications and Interoperable Systems

View full text Add to dashboard Cite

show abstract

“…Overlays are increasingly used as a fundamental building block of modern distributed systems, with numerous applications [5,8,11,13,15,22,25]. Unfortunately, many popular overlay construction protocols [1,10,27] do not usually take into account the underlying network infrastructure on which an overlay is deployed, and those that do tend to be limited to a narrow family of applications or overlays [29,30].…”

Section: Introductionmentioning

confidence: 99%

Fluidify: Decentralized Overlay Deployment in a Multi-cloud World

Resmi

Taı̈ani

2015

Distributed Applications and Interoperable Systems

View full text Add to dashboard Cite

Abstract. As overlays get deployed in large, heterogeneous systems-ofsystems with stringent performance constraints, their logical topology must exploit the locality present in the underlying physical network. In this paper, we propose a novel decentralized mechanism-Fluidify-for deploying an overlay network on top of a physical infrastructure while maximizing network locality. Fluidify uses a dual strategy that exploits both the logical links of an overlay and the physical topology of its underlying network. Simulation results show that in a network of 25,600 nodes, Fluidify is able to produce an overlay with links that are on average 94% shorter than that produced by a standard decentralized approach based on slicing, while demonstrating a sub-linear time complexity.

show abstract

“…Decentralized topology construction protocols [1], [2], [3] are natural candidates to support the numerous overlay networks that have been proposed for over a decade now in the context of many P2P and cloud-based applications such as telecommunications (Skype), streaming systems [4], [5], pubsub systems [6], [7], and key-value stores [3], [8], [9]. Among these, gossip topology construction protocols have received a large amount of attention [1], [2], [10], [11] due to their inherent ability to scale, survive, and adapt.…”

Section: Introductionmentioning

confidence: 99%

Polystyrene: the Decentralized Data Shape That Never Dies

Bouget

Kervadec

Kermarrec

et al. 2014

2014 IEEE 34th International Conference on Distributed Computing Systems

Self Cite

View full text Add to dashboard Cite

Abstract-Decentralized topology construction protocols organize nodes along a predefined topology (e.g. a torus, ring, or hypercube). Such topologies have been used in many contexts ranging from routing and storage systems, to publish-subscribe and event dissemination. Since most topologies assume no correlation between the physical location of nodes and their positions in the topology, they do not handle catastrophic failures well, in which a whole region of the topology disappears. When this occurs, the overall shape of the system typically gets lost. This is highly problematic in applications in which overlay nodes are used to map a virtual data space, be it for routing, indexing or storage. In this paper, we propose a novel decentralized approach that maintains the initial shape of the topology even if a large (consecutive) portion of the topology fails. Our approach relies on the dynamic decoupling between physical nodes and virtual ones enabling a fast reshaping. For instance, our results show that a 51,200-node torus converges back to a full torus in only 10 rounds after 50% of the nodes have crashed. Our protocol is both simple and flexible and provides a novel form of collective survivability that goes beyond the current state of the art.

show abstract

Heterogeneous Gossip

Cited by 22 publications

References 30 publications

Scalable Anti-KNN: Decentralized Computation of k-Furthest-Neighbor Graphs with HyFN

Scalable Anti-KNN: Decentralized Computation of k-Furthest-Neighbor Graphs with HyFN

Fluidify: Decentralized Overlay Deployment in a Multi-cloud World

Polystyrene: the Decentralized Data Shape That Never Dies

Contact Info

Product

Resources

About