Laurent Pellegrino

A Distributed Publish/Subscribe System for RDF Data

The pub/sub communication style is a prevalent messaging pattern for filtering information from distributed and large-scale network (e.g., from the real-time web, sensor networks, etc.) thanks to the decoupling between publishers and subscribers. At the same time, persisting the published information is a prerequisite for any further batch analytics on such big amount of data. As data can be heterogeneous, reliance on format from the semantic web such as RDF is unavoidable. In this paper we introduce two versions of a content-based pub/sub matching algorithm for RDF described events, working on an adapted version of the CAN structured P2P network designed to both store and disseminate RDF events. In contrary to existing pub/sub solutions based upon structured overlay networks that index semantic events several times due to the use of hash functions, we leverage the lexicographic order of the event elements. Thus, only subscriptions and not publications have to be duplicated, which is better given that in real settings, publications may occur more frequently than subscriptions. Furthermore, our system allows to publish events made of any number of elements and the subscription language leverages the SPARQL query language. The first algorithm we introduce initially derives from the ideas discussed by Liarou. et al. based upon rewriting continuous queries along matching RDF elements (CSBV) with the purpose to perform the matching between subscriptions and several RDF elements on multiple nodes. The experimental results discuss the applicability of the presented algorithms to some synthetic scenarios and identify, accordingly, which pub/sub matching algorithm is the more relevant.

A Generic API for Load Balancing in Structured P2P Systems

Real world datasets are known to be highly skewed, often leading to an important load imbalance issue for distributed systems managing them. To address this issue, there exist almost as many load balancing strategies as there are different systems. When designing a scalable distributed system geared towards handling large amounts of information, it is often not so easy to anticipate which kind of strategy will be the most efficient to maintain adequate performance regarding response time, scalability and reliability at any time. Based on this observation, we describe the methodology behind the building of a generic API to implement and experiment any strategy independently from the rest of the code, prior to a definitive choice for instance. We then show how this API is compatible with famous existing systems and their load balancing scheme. We also present results from our own distributed system which targets the continuous storage of events structured according to the Semantic Web standards, further retrieved by interested parties. As such, our system constitutes a typical example of a Big Data environment.

A generic API for load balancing in distributed systems for big data management

Distributed systems for big data management very often face the problem of load imbalance among nodes. To address this issue, there exist almost as many load balancing strategies as there are different systems. When designing a scalable distributed system geared towards handling large amounts of information, it is often not so easy to anticipate which kind of strategy will be the most efficient to maintain adequate performance regarding response time, scalability, and reliability at any time. Based on this observation, we describe a generic API to implement and experiment any strategy independently from the rest of the code, prior to a definitive choice for instance. We then show how existing load balancing strategies used by famous systems could be implemented with this API. We also present how this work has helped us implement load balancing on our distributed system and modify the behavior of our strategy in a few lines of code. This led us to easily perform various experiments to determine the most efficient scheme for our system. This paper is an extension to our work presented at Workshop on Parallel and Distributed Computing for Big Data Applications (WPBA) 2014. We detail here more experiments and extend the use of the API to a broad class of big data storage systems. Copyright

Efficient data-intensive event-driven interaction in SOA

This paper presents a middleware that enables the efficient delivery of events carrying large attachments. We transparently decouple event-description from event-data, in order to avoid useless data-transfers and modifications to endpoints business logic. Our solution relieves the event-delivery system of large data transfers, by enabling direct, but transparent, publisher to subscriber data-exchange. The experiments show that we can reduce the average event delivery time by half, compared to a standard approach requiring the full mediation of the event-delivery system.

Improving data-intensive EDA performance with annotation-driven laziness

Event-driven programming in large scale environments is becoming a common requirement of modern distributed applications. It introduces some beneficial effects such as real-time state updates and replications, which enable new kinds of applications and efficient architectural solutions. However, these benefits could be compromised if the adopted infrastructure were not designed to ensure efficient delivery of events and related data. This paper presents an architectural model, a middleware (WS-Link) and annotation-based mechanisms to ensure high performance in delivering events carrying large attachments. We transparently decouple event notification from related data to avoid useless data-transfers. This way, while event notifications are routed in a conventional manner through an event service, data are directly and transparently transferred from publishers to subscribers. The theoretical analysis shows that we can reduce the average event delivery time by half, compared to a conventional approach requiring the full mediation of the event service. The experimental analysis confirms that the proposed approach outperforms the conventional one (both for throughput and delivery time) even though the middleware overhead, introduced by the specific adopted model, slightly reduces the expected benefits. A novel architectural model for improving the efficiency of publish/subscribe middleware.A framework (WS-Link) that implements the model for SOA-based event-driven interactions.Java annotations to program the lazy strategies that guide the framework.An intense experimentation analysis of a complete implementation of the framework in a Grid5000-based testbed.