Section: New Results

Fault Management and Causal Analysis

Participants : Pascal Fradet, Alain Girault, Gregor Goessler, Jean-Bernard Stefani, Martin Vassor.

Fault Ascription in Concurrent Systems

The failure of one component may entail a cascade of failures in other components; several components may also fail independently. In such cases, elucidating the exact scenario that led to the failure is a complex and tedious task that requires significant expertise.

The notion of causality (did an event $e$ cause an event $e^{\text{'}}$?) has been studied in many disciplines, including philosophy, logic, statistics, and law. The definitions of causality studied in these disciplines usually amount to variants of the counterfactual test “$e$ is a cause of $e^{\text{'}}$ if both $e$ and $e^{\text{'}}$ have occurred, and in a world that is as close as possible to the actual world but where $e$ does not occur, $e^{\text{'}}$ does not occur either”. In computer science, almost all definitions of logical causality — including the landmark definition of [48] and its derivatives — rely on a causal model that. However, this model may not be known, for instance in presence of black-box components. For such systems, we have been developing a framework for blaming that helps us establish the causal relationship between component failures and system failures, given an observed system execution trace. The analysis is based on a formalization of counterfactual reasoning [6].

We are currently working on a revised version of our general semantic framework for fault ascription in  [46] that satisfies a set of formally stated requirements — such as its behavior under several notions of abstraction and refinement —, and on its instantiation to acyclic models of computation, in order to compare our approach with the standard definition of actual causality proposed by Halpern and Pearl.

Fault Management in Virtualized Networks

From a more applied point of view we are investigating, in the context of Sihem Cherrared's PhD thesis, approaches for fault explanation and localization in virtualized networks. In essence, Network Function Virtualization (NFV), widely adopted by the industry and the standardization bodies, is about running network functions as software workloads on commodity hardware to optimize deployment costs and simplify the life-cycle management of network functions. However, it introduces new fault management challenges including dynamic topology and multi-tenant fault isolation that we discuss in [14]. As a first step to tackle those challenges, we have extended the classical fault management process to the virtualized functions by introducing LUMEN: a Global Fault Management Framework. Our approach aims at providing the availability and reliability of the virtualized 5G end-to-end service chain. LUMEN includes the canonical steps of the fault management process and proposes a monitoring solution for all types of Network virtualization Environments. Our framework is based on open source solutions and could easily be integrated with other existing autonomic management models.