Figure 4 - uploaded by Tal Anker
Content may be subject to copyright.
Source publication
. Group communication is a useful abstraction in the development of highly available distributed and communication-oriented applications in wide area networks (WANs). The most important aspects of this abstraction are the dynamic maintenance of group membership and its diverse semantics for interleaving membership change notifications within the fl...
Context in source publication
Context 1
... 5.2. Consider Example 4.1 (illustrated in Figure 4) where two pro- cesses, c and d, concurrently try to join group G. Assume that the current view of group G is < G; fa; bg; 3 >. Consider the case in which membership server M 1 rst receives the notiication that c is joining, and then issues proposals for the view < G; fa; b; cg;fg;4 > to a and c. ...
Similar publications
Purpose
The purpose of this paper is to provide a mechanism to characterize web sites, at the conceptualization stage, whose information content can be defined in terms of hybrid conceptual models. Usually in this type of web site, some of the domain information units are modeled in terms of classes/entities while others remain as heterogeneous irr...
Citations
... he problem was widely simulated at the study community. he reason for this great simulation is that many distributed systems need a group membership protocol [3][4][5][6][7] . In spite of such practically usefulness, to our knowledge there is only a few labor that have been committed to this problem in a mobile ad hoc computing environment. ...
In distributed systems, a group of computer should continue to do cooperation in order to finish some jobs. In such a system, a group membership protocol is especially practical and important elements to provide processes in a group with a consistent common knowledge of the membership of the group. Whenever a membership change occurs, processes should agree on which of them should do to accomplish an unfinished job or begins a new job. The problem of knowing a stable membership view is very same with the one of agreeing common predicate in a distributed system such as the consensus problem. Based on the stopping investigating protocol that is traditional one in asynchronous distributed systems, we present the new group membership protocol in mobile wireless networks.
... There exist other communication's models for distributed systems, such as BPCE. For example, following the Virtual Synchrony [16], the Extended Virtual Synchrony [17] was proposed, and, more recently, works in this area focuses on scaling group membership [18], [19]. To address this third problem, we implemented a synchronization mechanism for RO's among peers, inspired on the Virtual Synchrony [16]. ...
Business process management systems are largely used nowadays. Process Modeling, whether for depicting new processes or for optimizing existing ones, requires knowledge and proper tools. However, most process models are created from scratch, not having reuse promoted, i.e., not considering existing knowledge. Large enterprises not using a unique integrated system, and also some of them that do, have the same business process implemented in a variety of ways, due to differences in their units, culture or environments. In this work we evaluate BPCE, a P2P tool we previously proposed as a way of collaboratively developing process models. We describe tool implementation issues, a modeling case study, a simulation conducted using the tool and the changes in the original proposal caused by those investigations. The results present evidence that such a tool can minimize the time to develop new models, reduce the differences among similar business processes conducted in distinct organization units, enhance the quality of process design and promote reuse.
... More alarmingly, the rate of member join and leave events will likely grow linearly with the number of members as well, possibly leading to an unmanageable amount of network bandwidth or latency. Many of these concerns have been addressed previously [2] [14] [15] [28]. We believe Fireflies can scale easily to thousands of members and that this is sufficient for many applications. ...
... Onehop peer-to-peer routing protocols that maintain full membership for fail-stop environments are presented in [14] [15]. CONGRESS [2] is a non-P2P solution based on a scalable server hierarchy. Neither tolerates Byzantine behavior. ...
... More alarmingly, the rate of member join and leave events will likely grow linearly with the number of members as well, possibly leading to an unmanageable amount of network bandwidth or latency. Many of these concerns have been addressed previously [2, 14, 15, 28]. We believe Fireflies can scale easily to thousands of members and that this is sufficient for many applications. ...
This paper describes and evaluates Fireflies, a scalable pro- tocol for supporting intrusion-tolerant network overlays.1 While such a protocol cannot distinguish Byzantine nodes from correct nodes in general, Fireflies provides correct nodes with a reasonably current view of which nodes are live, as well as a pseudo-random mesh for communication. The amount of data sent by correct nodes grows linearly with the aggregate rate of failures and recoveries, even if pro- voked by Byzantine nodes. The set of correct nodes form a connected submesh; correct nodes cannot be eclipsed by Byzantine nodes. Fireflies is deployed and evaluated on PlanetLab.
... The authors of Congress [6] and Maestro [24] were the rst to observe that separating maintenance of membership from a group multicast will better enhance scalability of fault-tolerant distributed applications . This separation was later adopted by researchers who addressed the WAN environment ([63, 7, 93]). InterGroup [21] presents another WAN approach to address scalability. ...
... @BULLET The design demonstrates how to more effectively decouple the algorithm for achieving Virtual Synchrony from the algorithm for maintaining membership. As suggested in [9, 58], such efficient decoupling is important for providing scalable GCSs in WANs. ...
... Efficient decoupling of membership and Virtually Synchronous multicast algorithms allows for an architecture in which the membership service is implemented by a small set of dedicated membership servers maintaining the membership information on behalf of a large set of clients. This architecture was proposed in [9, 58] for supporting scalable membership services in WANs. Our work extends this architecture by specifying how it can be used as a base for a Virtually Synchronous GCS. ...
... The new architecture decouples two key components of GCSs — Virtual Synchrony and Membership — allowing them to execute asynchronously and in parallel. Such decoupling has been regarded as critical for providing scalable and efficient group communication services in wide-area networks: It allows for the utilization of a scalable architecture for group membership in which a small set of membership servers maintains group membership of a large set of clients [9]. The architecture suggested here has been adopted by the Moshe system developed at UCSD and MIT [58] . ...
Group communication services, and especially those that implement Virtual Synchrony semantics, are powerful middleware systems that facilitate the development of fault-tolerant distributed applications. In this thesis, we present a high quality, theoretical design of a group communication service that implements Virtual Synchrony semantics and is aimed for deployment in wide-area networks (WANs). The design features a novel algorithm for implementing Virtual Synchrony semantics; the algorithm is more appropriate for WANs than the existing solutions because it involves fewer rounds of communication and operates in a scalable WAN-oriented architecture. The high quality of the design refers to the level of formality and rigor at which it is done: The design includes formal and precise specifications, algorithms, correctness proofs, and performance analyses. We develop the necessary supporting theory and methodology required for producing and evaluating this design. In particular, we develop a formal, inheritance-based, methodology that supports incremental construction of specifications, models, and proofs. This methodology helps us manage the complexity of the design and makes it evident which part of the algorithm implements which property of the system. We also develop new, formal approaches in the area of performance evaluation. Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2002. Includes bibliographical references (p. 159-165).
... These systems explored several different models of group communication such as VS [3] and EVS [4]. More recent work in this area focuses on scaling group membership to wide area networks (WAN) [21], [22]. With the increased use of GCS's over insecure open networks , some research interests shifted to securing these systems . ...
Group communication systems are high-availability distributed systems providing reliable and ordered message delivery, as well as a membership service, to group-oriented applications. Many such systems are built using a distributed client-server architecture where a relatively small set of servers provide service to numerous clients. In this work, we show how group communication systems can be enhanced with security services without sacrificing robustness and performance. More specifically, we propose several integrated security architectures for distributed client-server group communication systems. In an integrated architecture, security services are implemented in servers, in contrast to a layered architecture, where the same services are implemented in clients. We discuss performance and accompanying trust issues of each proposed architecture and present experimental results that demonstrate the superior scalability of an integrated architecture.
... These systems explored several different models of group communication such as Virtual Synchrony [4] and Extended Virtual Synchrony [12]. More recent work in this area focuses on scaling group membership to wide-area networks [42], [43]. Research on securing group communication is fairly new. ...
Contributory group key agreement protocols generate group keys based on contributions of all group members. Particularly appropriate for relatively small collaborative peer groups, these protocols are resilient to many types of attacks. Unlike most group key distribution protocols, contributory group key agreement protocols offer strong security properties such as key independence and perfect forward secrecy. We present the first robust contributory key agreement protocol resilient to any sequence of group changes. The protocol, based on the Group Diffie-Hellman contributory key agreement, uses the services of a group communication system supporting virtual synchrony semantics. We prove that it provides both virtual synchrony and the security properties of Group Diffie-Hellman, in the presence of any sequence of (potentially cascading) node failures, recoveries, network partitions, and heals. We implemented a secure group communication service, Secure Spread, based on our robust key agreement protocol and Spread group communication system. To illustrate its practicality, we compare the costs of establishing a secure group with the proposed protocol and a protocol based on centralized group key management, adapted to offer equivalent security properties.
... Our understanding of scalable membership protocol should not be confused with that of [1], [7] where the aim is to provide each member of the group with an accurate and timely global view of the membership. The problem we consider is instead to provide each node with partial membership information which is sufficient to achieve reliable dissemination using a traditional gossip-based protocol. ...
Gossip-based protocols have received considerable attention for broadcast applications due to their attractive scalability and reliability properties. The reliability of probabilistic gossip schemes studied so far depends on each user having knowledge of the global membership and choosing gossip targets uniformly at random. The requirement of global knowledge is undesirable in large-scale distributed systems.
In this paper, we present a novel peer-to-peer membership service which operates in a completely decentralized manner in that nobody has global knowledge of membership. However, membership information is replicated robustly enough to support gossip with high reliability. Our scheme is completely self-organizing in the sense that the size of local views naturally converges to the ‘right’ value for gossip to succeed. This ‘right’ value is a function of system size, but is achieved without any node having to know the system size. We present the design, theoretical analysis and preliminary evaluation of Scamp. Simulations show that its performance is comparable to that of previous schemes which use global knowledge of membership at each node.
... Though these approaches have proven scalable for message dissemination, they rely on a nonscalable membership protocol: 1 They assume that the subset of nodes that a node gossips to is chosen uniformly among all participating nodes, requiring that each node should know every other node. This imposes high requirements on memory and synchronization, which adversely affects their scalability. ...
... As the number of participating nodes changes, we show both analytically and through simulation that the size of partial views automatically adapts to the desired value. These results are 1. Our understanding of scalable membership protocol should not be confused with that of [1], [16], where the aim is to provide each member of the group with an accurate and timely global view of the membership. ...
... These results are 1. Our understanding of scalable membership protocol should not be confused with that of [1], [16], where the aim is to provide each member of the group with an accurate and timely global view of the membership. The problem we consider is instead to achieve reliable multicast without requiring global knowledge of membership at each node. ...
Gossip-based 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 large-scale groups. In this paper, we present SCAMP (Scalable Membership protocol), a novel peer-to-peer membership protocol which operates in a fully decentralized manner and provides each member with a partial view of the group membership. Our protocol is self-organizing 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.
... Moshe is implemented as part of a novel group membership service for computer supported cooperative work (CSCW) applications in WANs [Anker et al. 1998]. Moshe is complemented by a virtually synchronous communication service [Keidar and Khazan 2000], and it is partitionable Chockler et al. 2001;Babaoglu et al. 2001], that is, several disjoint views can exist concurrently. ...
... Moshe is part of a novel architecture for group membership services designed for CSCW applications in WANs [Anker et al. 1998]. This architecture employs a client-server approach: group membership services are provided by dedicated membership servers, which themselves are not members of any multicast group. ...
... (2) The notification service behaves like an eventually perfect failure detector (see [Chockler et al. 2001;Babaoglu et al. 2001]), that is, it eventually stops making mistakes. A similar guarantee is formally defined in terms of network stability and failure detector properties in [Anker et al. 1998;Chockler et al. 2001]. ...
We present Moshe, a novel scalable group membership algorithm built specifically for use in wide area networks (WANs), which can suffer partitions. Moshe is designed with three new significant features that are important in this setting: it avoids delivering views that reflect out-of-date memberships; it requires a single round of messages in the common case; and it employs a client-server design for scalability. Furthermore, Moshe's interface supplies the hooks needed to provide clients with full virtual synchrony semantics. We have implemented Moshe on top of a network event mechanism also designed specifically for use in a WAN. In addition to specifying the properties of the algorithm and proving that this specification is met, we provide empirical results of an implementation of Moshe running over the Internet. The empirical results justify the assumptions made by our design and exhibit good performance. In particular, Moshe terminates within a single communication round over 98% of the time. The experimental results also lead to interesting observations regarding the performance of membership algorithms over the Internet.
