Results 1 
8 of
8
Peertopeer membership management for gossipbased protocols
 IEEE TRANSACTIONS ON COMPUTERS
, 2003
"... Gossipbased protocols for group communication have attractive scalability and reliability properties. The probabilistic gossip schemes studied so far typically assume that each group member has full knowledge of the global membership and chooses gossip targets uniformly at random. The requirement ..."
Abstract

Cited by 219 (23 self)
 Add to MetaCart
(Show Context)
Gossipbased protocols for group communication have attractive scalability and reliability properties. The probabilistic gossip schemes studied so far typically assume that each group member has full knowledge of the global membership and chooses gossip targets uniformly at random. The requirement of global knowledge impairs their applicability to very largescale groups. In this paper, we present SCAMP (Scalable Membership protocol), a novel peertopeer membership protocol which operates in a fully decentralized manner and provides each member with a partial view of the group membership. Our protocol is selforganizing in the sense that the size of partial views naturally converges to the value required to support a gossip algorithm reliably. This value is a function of the group size, but is achieved without any node knowing the group size. We propose additional mechanisms to achieve balanced view sizes even with highly unbalanced subscription patterns. We present the design, theoretical analysis, and a detailed evaluation of the basic protocol and its refinements. Simulation results show that the reliability guarantees provided by SCAMP are comparable to previous schemes based on global knowledge. The scale of the experiments attests to the scalability of the protocol.
The Peer Sampling Service: Experimental Evaluation of Unstructured GossipBased Implementations
 In Middleware ’04: Proceedings of the 5th ACM/IFIP/USENIX international conference on Middleware
, 2004
"... Abstract. In recent years, the gossipbased communication model in largescale distributed systems has become a general paradigm with important applications which include information dissemination, aggregation, overlay topology management and synchronization. At the heart of all of these protocols l ..."
Abstract

Cited by 184 (42 self)
 Add to MetaCart
(Show Context)
Abstract. In recent years, the gossipbased communication model in largescale distributed systems has become a general paradigm with important applications which include information dissemination, aggregation, overlay topology management and synchronization. At the heart of all of these protocols lies a fundamental distributed abstraction: the peer sampling service. In short, the aim of this service is to provide every node with peers to exchange information with. Analytical studies reveal a high reliability and efficiency of gossipbased protocols, under the (often implicit) assumption that the peers to send gossip messages to are selected uniformly at random from the set of all nodes. In practice—instead of requiring all nodes to know all the peer nodes so that a random sample could be drawn—a scalable and efficient way to implement the peer sampling service is by constructing and maintaining dynamic unstructured overlays through gossiping membership information itself. This paper presents a generic framework to implement reliable and efficient peer sampling services. The framework generalizes existing approaches and makes it easy to introduce new ones. We use this framework to explore and compare several implementations of our abstract scheme. Through extensive experimental analysis, we show that all of them lead to different peer sampling services none of which is uniformly random. This clearly renders traditional theoretical approaches invalid, when the underlying peer sampling service is based on a gossipbased scheme. Our observations also help explain important differences between design choices of peer sampling algorithms, and how these impact the reliability of the corresponding service. 1
Parsimonious flooding in dynamic graphs
 In Proc. of 28th Symp. on Principles of Distributed Computing (PODC
, 2009
"... An edgeMarkovian process with birthrate p and deathrate q generates sequences of graphs (G0, G1, G2,...) with the same node set [n] such that Gt is obtained from Gt−1 as follows: if e / ∈ E(Gt−1) then e ∈ E(Gt) with probability p, and if e ∈ E(Gt−1) then e / ∈ E(Gt) with probability q. Clementi e ..."
Abstract

Cited by 38 (1 self)
 Add to MetaCart
(Show Context)
An edgeMarkovian process with birthrate p and deathrate q generates sequences of graphs (G0, G1, G2,...) with the same node set [n] such that Gt is obtained from Gt−1 as follows: if e / ∈ E(Gt−1) then e ∈ E(Gt) with probability p, and if e ∈ E(Gt−1) then e / ∈ E(Gt) with probability q. Clementi et al. (PODC 2008) analyzed thoroughly information dissemination in such dynamic graphs, by establishing bounds on their flooding time — flooding is the basic mechanism in which every node becoming aware of an information at step t forwards this information to all its neighbors at all forthcoming steps t ′> t. In this paper, we establish tight bounds on the complexity of flooding for all possible birth rates and death rates, completing the previous results by Clementi et al. Moreover, we note that despite its many advantages in term of simplicity and robustness, flooding suffers from its high bandwidth consumption. Hence we al! so show that flooding in dynamic graphs can be implemented in a more parsimonious manner, so that to save bandwidth, yet preserving efficiency in term of simplicity and completion time. For a positive integer k, we say that the flooding protocol is kactive if each node forwards an information only during the k time steps immediately following the step at which the node receives that information for the first time. We define the reachability threshold for the flooding protocol as the smallest integer k such that, for any source s ∈ [n], the kactive flooding protocol from s completes (i.e., reaches all nodes), and we establish tight bounds for this parameter. We show that, for a large spectrum of parameters p and q, the reachability threshold is by several orders of magnitude smaller than the flooding time. In particular, we show A part of this work was done during the stay of the second
R.S.: Performance of an intrusiontolerant gossip protocol
 In: Proc. of the 21st IASTED International Conference on Parallel and Distributed Computing and Systems (PDCS
, 2009
"... Gossipbased communication protocols have a number of desirable properties, including scalability and resilience to crash faults. However, very little work has been done in developing gossip protocols that are intrusiontolerant, i.e., that are able to provide useful services even while the system ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
Gossipbased communication protocols have a number of desirable properties, including scalability and resilience to crash faults. However, very little work has been done in developing gossip protocols that are intrusiontolerant, i.e., that are able to provide useful services even while the system is under attack or has been compromised. We present implementation details and performance results for StarblabIT, an intrusiontolerant gossipbased group communication system. Our system combines the scalability and resilience of a gossip protocol along with the ability to provide useful services even during and after the system has been compromised or damaged due to an accident or intrusion. Performance results for our system indicate that the latency is scalable to large systems and that the throughput is quite high for an intrusiontolerant system.
Implementation and performance testing of a gossipbased communication system
 in Proceedings of the 19th IASTED International Conference on Parallel and Distributed Computing and Systems
, 2007
"... Computer systems today are increasingly complex, employing vast numbers of nodes connected over large geographic areas. As these types of systems become more pervasive and integral to a variety of services, there is a heightened need for scalability and reliability in the underlying communication ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
(Show Context)
Computer systems today are increasingly complex, employing vast numbers of nodes connected over large geographic areas. As these types of systems become more pervasive and integral to a variety of services, there is a heightened need for scalability and reliability in the underlying communication system. We present an implementation of a gossipbased, widearea group communication system that is resilient to process and communication failures and is scalable to a very large number of nodes. We provide implementation details and describe our performance tests. The results are presented and analyzed.
Intrusiontolerant dissemination in largescale systems
 In Proceedings of the 2008 International Conference on Parallel and Distributed Processing Techniques and Applications, pages 934–940, Las Vegas, NV
, 2008
"... Abstract With the growth of the Internet and increased demand for Web Services has come a heightened need for scalable widearea group communication systems. The need for trust in such systems and the vulnerability of such systems to attack provide motivation to develop protocols that are intrusi ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Abstract With the growth of the Internet and increased demand for Web Services has come a heightened need for scalable widearea group communication systems. The need for trust in such systems and the vulnerability of such systems to attack provide motivation to develop protocols that are intrusiontolerant and able to provide critical services even during an attack or in the presence of arbitrary faults, errors, and accidents. We present StarblabIT, a new gossipbased group communication system that is scalable and resilient to process and communication failures, including arbitrary (Byzantine) faults and malicious attack. We employ a modular approach to the design, and build on prior work that generalizes the transformation of crashtolerant protocols to Byzantinetolerant protocols. The StarblabIT protocol avoids the need for signatures on most messages by making use of message digests embedded in a single signed message to achieve efficient and secure message delivery.
The fi...
"... Abstract. This paper considers the quasirandom rumor spreading model introduced by Doerr, Friedrich, and Sauerwald in [SODA 2008], hereafter referred to as the listbased model. Each node is provided with a cyclic list of all its neighbors, chooses a random position in its list, and from then on ca ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract. This paper considers the quasirandom rumor spreading model introduced by Doerr, Friedrich, and Sauerwald in [SODA 2008], hereafter referred to as the listbased model. Each node is provided with a cyclic list of all its neighbors, chooses a random position in its list, and from then on calls its neighbors in the order of the list. This model is known to perform asymptotically at least as well as the random phonecall model, for many network classes. Motivated by potential applications of the listbased model to live streaming, we are interested in its worst case behavior. Our first main result is the design of an O(m + n log n)time algorithm that, given any nnode medge network G, and any sourcetarget pair s, t ∈ V (G), computes the maximum number of rounds it may take for a rumor to be broadcast from s to t in G, in the listbased model. This algorithm yields an O(n(m + n log n))time algorithm that, given any network G, computes the maximum number of rounds it may take for a rumor to be broadcast from any source to any target, in the listbased model. Hence, the listbased model is computationally easy to tackle in its basic version. The situation is radically different when one is considering variants of the model in which nodes are aware of the status of their neighbors, i.e., are aware of whether or not they have already received the rumor, at any point in time. Indeed, our second main result states that, unless P = NP, the worst case behavior of the listbased model with the additional feature that every node is perpetually aware of which of its neighbors have already received the rumor cannot be approximated in polynomial time within a ( 1 n) 1 2 −ɛ multiplicative factor, for any ɛ> 0. As a byproduct of this latter result, we can show that, unless P = NP, there are no PTAS enabling to approximate the worst case behavior of the listbased model, whenever every node perpetually keeps track of the subset of its neighbors which have sent the rumor to it so far.