Gossip Messages Research Articles

Scalable epidemic message passing interface fault tolerance

Resilience and fault tolerance are challenging tasks in the field of high performance computing (HPC) and extreme scale systems. Components fail more often in such systems, results in application abort. Adopting fault–tolerance techniques can be consistently detect failures and continue application’s execution even if the failures exist. A prominent parallel programming specification, message passing interface (MPI), as it would be used to implement failure detection and consensus algorithm in this paper. Although the MPI does not facilitate fault tolerant behavior, this work presents a fault tolerant, matrix based failure detection and consensus algorithm. The proposed algorithm uses Gossiping. To detect failures, randomised pinging will be applied during the execution of the algorithm by using piggybacked gossip messages. In order to achieve consensus on the failures in the system, failed processes’ information will be sent using the same piggybacked gossip messages to all the alive processes. The algorithm was implemented in MPI framework and is completely fault tolerant. The results exhibit all the MPI process failures were detected using randomised pinging and global consensus has achieved on failed MPI process in the system.

Restrictive deterrence and the scope of hackers’ reoffending: Findings from two randomized field trials

Extensive criminological research has investigated the effectiveness of law enforcement efforts to prevent offenders' reoffending and effect changes in criminal behavior patterns. However, no research has quantified the influence of gossip regarding law enforcement operations and its effect on repeat offending. Moreover, only scant research has studied these relationships in the cyber-environment. The current work addresses this gap by studying the effect of online messages sent to active hackers that describe law enforcement efforts against website defacement activity. In two field experiments, we randomly assigned self-identified hackers with active Facebook accounts into control and treatment groups. We then sent subjects in the treatment group a gossip message (via Facebook) alerting them to law enforcement efforts in cyberspace. Following the intervention, we compared the control and treatment groups in terms of changes in the proportion of reoffending, number of website defacements, and severity of website defacements. Findings reveal that when the gossip was sent to a hacker's private inbox, it prompted a reduction in the proportion of hackers who reoffend, the frequency at which they reoffend, and the severity of attacks they generate. However, posting similar gossip on the hacker's Facebook wall was ineffective in restricting malicious hacking activity. Theoretical implications and practical recommendations for law enforcement operations in cyberspace are discussed.

Reinforced Secure Gossiping Against DoS Attacks in Post-Disaster Scenarios

During and after a disaster, the perceived quality of communication networks often becomes remarkably degraded with an increased ratio of packet losses due to physical damages of the networking equipment, disturbance to the radio frequency signals, continuous reconfiguration of the routing tables, or sudden spikes of the network traffic, e.g., caused by the increased user activity in a post-disaster period. Several techniques have been introduced so far (mainly using data retransmission mechanisms) to tolerate such circumstances. Among them, gossiping has been shown to be efficient in the recovery from message losses. However, a conventional gossiping scheme may exhibit security problems, which can be exploited for further attacks (such as Denial of Service - DoS attack). For instance, the flooding method used by the gossiping can be used to forward the traffic towards many vulnerable nodes to drain their resources and compromise them. Typically, protection against DoS attacks is realized by using cryptographic primitives. However, their scalability limits and costs make them improper for emergency communications after a disaster. In this article, we introduce an approach based on reinforcement learning and game theory to protect the gossiping scheme from DoS attacks without incurring the costs of cryptographic primitives. In our method, nodes properly select which requests to satisfy, which in turn helps other nodes to avoid receiving manipulated gossip messages from malicious and colluded nodes. Additionally, our method operates without exploiting any cryptographic primitives, which prevents excessive energy waste that is undesired in post-disaster resilient networking. Simulation experiments performed in OMNeT++ confirmed the advantages of our approach over the reference schemes in terms of reliability, security, overhead, latency, and power efficiency.

Improving the gossiping effectiveness with distributed strategic learning (Invited paper)

Gossiping is a widely known and successful approach to reliable communications, tolerating packet losses and link crashes. It has been extensively used in several middleware kinds, such as event notification services and application domains, like infrastructures for air traffic management, power grid control, health information exchange, just to cite some of them. Despite achieving a high loss-tolerance and scalability degrees, gossiping is affected by degraded performances and heavy traffic loads on the network. For this reason, it may be not optimal in applications where reliability must be provided jointly with timeliness and/or in congestion-prone networks. The crucial aspect for improving a gossiping scheme is deciding which nodes should receive a gossiping message, and our driving idea is to adopt a distributed strategic learning logic to determine such nodes in an efficient manner. This is able to resolve gossiping’s weakness points and to achieve better performance and reduced traffic loads.This paper describes how to introduced strategic learning in a gossip scheme so as to determine the best set of nodes that can be used to send gossip messages and to optimize their utility. Such a solution has been experimentally assessed through a set of simulations demonstrating the effectiveness of the proposal.

ReCREW: A Reliable Flash-Dissemination System

In this paper, we explore a new form of dissemination that arises in distributed, mission-critical applications called Flash Dissemination. This involves the rapid dissemination of rich information to a large number of recipients in a very short period of time. A key characteristic of Flash Dissemination is its unpredictability (e.g., natural hazards), but when invoked it must harness all possible resources to ensure timely delivery of information. Additionally, it must scale to a large number of recipients and perform efficiently in highly heterogeneous (data, network) and failure prone environments. We investigate a peer-based approach based on the simple principle of transferring dissemination load to information receivers using foundations from broadcast networks, gossip theory, and random networks. Gossip-based protocols are well known for being stateless, scalable, and fault-tolerant; however, their performance degrades as content size increases, because of the propagation of redundant gossip messages. In this paper, we propose Concurrent Random Expanding Walkers (CREW), a smart gossip protocol designed to maximize the speed of dissemination by transmitting data only as needed, and by exploiting both intra- and internode concurrency. CREW is designed to support both content and network heterogeneity and deal with transmission failures without sacrificing dissemination speed. We implemented CREW on top of a scalable middleware environment that allows for deployment across several platforms and developed optimizations without compromising on the stateless nature of CREW. We evaluated CREW empirically and compared it to optimized implementations of popular gossip and peer-based systems. Our experiments show that CREW significantly outperforms both traditional gossip and current large content dissemination systems while sustaining its performance in the presence of network errors.

Ramos: Concurrent writing and reconfiguration for collaborative systems

Collaborative systems, with specific distributed systems allow multiple participants to work in a common virtual space, while reproducing the different ways to interact in a group. Such systems have to manage not only the sharing of context and particularly the context consistency, but also at the same time the fault tolerance. No system in the literature combines these two requirements.In this paper, we are proposing the new protocol Ramos which implements a fault-tolerant, and a context consistency (ensuring a total order of write operations) based on an asynchronous message-passing model. Communication takes place via gossip messages, which are sent at any frequency between a dynamic set of nodes. Ramos is based on the Rambo III algorithm for replicated data services. Rambo III provides two functions: reconfiguration of a dynamic set of nodes and reading/writing of a replicated object. In Ramos the reconfiguration process from Rambo III is adapted to the needs of collaborative systems and Paxos is used to execute concurrent write operations. It is assumed that from a total set of 2f+1 nodes, at most a subset of f nodes is allowed to fail simultaneously. Furthermore, it is assumed that the application using Ramos provides a leader-election method. Ramos, the algorithm proposed here, provides one significant feature: all write operations are totally ordered.

Node discovery in networks

This paper abstracts the problem of network nodes discovering one another in a network of unknown size using all-to-all gossip. The problem is studied in terms of evolving directed graphs where each vertex represents a participating node and each edge represents one node’s knowledge about another. Ideally, such a graph has diameter one, i.e., each node knows all others. Nodes share their knowledge by sending gossip messages. Gossip among the nodes allows them to discover one another, decreasing the diameter of the graph. Here this problem is considered in several synchronous settings under different assumptions about the ability of the participating nodes to communicate. Specifically, the following aspects of communication are considered: (1) the ability of the nodes to multicast gossip messages, and (2) the size of the messages. The results describe the lower and upper bounds on the number of synchronous rounds required for the participants to discover each other. A particular question of interest is: if the network size is unknown, how does a node know that it has discovered all other nodes? Given a weakly-connected graph describing the initial knowledge of the nodes, every node in our algorithm can stop the discovery process knowing that there are no unknown nodes—this is done without any prior knowledge of the total number of nodes participating in the computation.

Long-lived Rambo: Trading knowledge for communication

Shareable data services providing consistency guarantees, such as atomicity (linearizability), make building distributed systems easier. However, combining linearizability with efficiency in practical algorithms is difficult. A reconfigurable linearizable data service, called Rambo, was developed by Lynch and Shvartsman. This service guarantees consistency under dynamic conditions involving asynchrony, message loss, node crashes, and new node arrivals. The specification of the original algorithm is given at an abstract level aimed at concise presentation and formal reasoning about correctness. The algorithm propagates information by means of gossip messages. If the service is in use for a long time, the size and the number of gossip messages may grow without bound. This paper presents a consistent data service for long-lived objects that improves on Rambo in two ways: it includes an incremental communication protocol and a leave service. The new protocol takes advantage of the local knowledge, and carefully manages the size of messages by removing redundant information, while the leave service allows the nodes to leave the system gracefully. The new algorithm is formally proved correct by forward simulation using levels of abstraction. An experimental implementation of the system was developed for networks-of-workstations. The paper also includes selected analytical and preliminary empirical results that illustrate the advantages of the new algorithm.

Active and passive techniques for group size estimation in large-scale and dynamic distributed systems

This paper presents two solutions to a distributed statistic collection problem, called Group Size Estimation. These algorithms are intended for large-scale and dynamic distributed systems such as Grids, peer-to-peer overlays, etc. Each algorithm estimates (both in a one-shot and continuous manner) the number of non-faulty processes present in the global group. The first active scheme samples receipt times of gossip messages, while the second passive scheme calculates the density of process identifiers when hashed to a real interval. Our analysis, trace-driven simulation and deployment on a 33-node Linux cluster study and compare the latencies, scalability, and accuracy of these schemes.

Gossip-Style Failure Detection and Distributed Consensus for Scalable Heterogeneous Clusters

Gossip protocols provide a means by which failures can be detected in large, distributed systems in an asynchronous manner without the limits associated with reliable multicasting for group communications. However, in order to be effective with application recovery and reconfiguration, these protocols require mechanisms by which failures can be detected with system-wide consensus in a scalable fashion. This paper presents three new gossip-style protocols supported by a novel algorithm to achieve consensus in scalable, heterogeneous clusters. The round-robin protocol improves on basic randomized gossiping by distributing gossip messages in a deterministic order that optimizes bandwidth consumption. Redundant gossiping is completely eliminated in the binary round-robin protocol, and the round-robin with sequence check protocol is a useful extension that yields efficient detection times without the need for system-specific optimization. The distributed consensus algorithm works with these gossip protocols to achieve agreement among the operable nodes in the cluster on the state of the system featuring either a flat or a layered design. The various protocols are simulated and evaluated in terms of consensus time and scalability using a high-fidelity, fault-injection model for distributed systems comprised of clusters of workstations connected by high-performance networks.

Two phase gossip: Managing distributed event histories

We describe a distributed protocol that operates on replicated data objects—of arbitrary abstract types—represented in terms of the history of object events, rather than in terms of object values. In general, a site that stores a replica of an object has only partial knowledge of the object's event history. We call such a replica an object representative. The goal of the protocol is to limit the sizes of the histories by checkpointing and discarding old events. We treat separately the three functions of (1) propagating events to sites that do not know about them, (2) checkpointing, and (3) discarding old events. A site rolls forward its checkpoint state as far as it knows its local version of the object's history to be complete, but it discards old events only as far as it knows that all other sites know their local histories to be complete. Our protocol propagates events among sites in gossip messages exchanged in the background in a two phase manner. Each site maintains several timestamp vectors indicating the extent of its knowledge of other sites' knowledge of events. These timestamp vectors are included in gossip messages and used to decide the extent of global completeness of each site's local version of the event history. We formally define the correctness criteria for our protocol in terms of the completeness properties of histories. The protocol is fault-tolerant in that site and communication failures can only slow down its progress. The communication and processing required by our protocol are done in the background at times arbitrarily chosen by each site. Thus, the system need not quiesce for the activities of checkpointing and discarding old events to proceed.