GUARDS: a generic upgradable architecture for real-time dependable systems

Abstract

Abstract. The development and validation of fault-tolerant computers for critical realtime applications are currently both costly and time-consuming. Often, the underlying technology is out-of-date by the time the computers are ready for deployment. Obsolescence can become a chronic problem when the systems in which they are embedded have lifetimes of several decades. This paper gives an overview of the work carried out in a project that is tackling the issues of cost and rapid obsolescence by defining a generic fault-tolerant computer architecture based essentially on commercial off-the-shelf (COTS) components (both processor hardware boards and real-time operating systems). The architecture uses a limited number of specific, but generic, hardware and software components to implement an architecture that can be configured along three dimensions: redundant channels, redundant lanes and integrity levels. The two dimensions of physical redundancy allow the definition of a wide variety of instances with different fault-tolerance strategies. The integrity level dimension allows application components of different levels of criticality to co-exist in the same instance. The paper describes the main concepts of the architecture, the supporting environments for development and validation, and the prototypes currently being implemented.

Reconciling fault-tolerant distributed computing and systems-on-chip

“…Well-known examples are MAFT [38], SAFEBUS [33], GUARDS [62] and TTP [39]. However, in sharp contrast to our problem, these systems incorporate hardware assistance only.…”

“…1, the basic idea of DARTS is to replace the common quartz oscillator and the clock tree by a fully distributed GALS-like approach [8] [72,78]). It has been proved elsewhere [62] that even such loose synchrony suffices for implementing metastability-free high-speed communication between different Fu's using bounded-size buffers. DARTS clocks (patented in [69]) provide a number of additional advantages, which makes them particularly promising for critical applications in the aerospace domain: First of all, the approach entirely circumvents quartz oscillators, which are fairly big and sensitive devices (shock, vibration, temperature etc.…”

VTV - A Voting Strategy for Real-Time Systems

“…In order to accomplish this, inexact voting strategies have been proposed [13,21,24]. This phenomenon is also observed in time domain due to several factors, such as clock drifts, node failures, processing and scheduling variations at node level, as well as communication delays.…”

Contribución a la validación de sistemas complejos tolerantes a fallos en la fase de diseño. Nuevos modelos de fallos y técnicas de inyección de fallos.

“…En efecto, con el fin de conseguir un máximo de confianza en el proceso de validación, es necesario aplicar conjuntamente varios métodos: la utilización aislada de un método no es suficiente para asegurar un buen nivel de Confiabilidad. En relación con esto, cabe citar las metodologías de concepción integrando validación formal, modelado analítico e inyección de fallos utilizadas en los proyectos SIFT [Schwartz y Melliar-Smith 1983], EVE [Arlat et al 1984], DELTA_4 [Powell 1988], IPDS (1 y 2) [Randell et al 1995], HIDE [Majzik y Bondavalli 1998], GUARDS [Powell et al 1999, Powell 2001 y DBench [DBench 2003]. En la actualidad se está observando una tendencia a utilizar las técnicas de validación formal en la validación de las especificaciones (como ocurre en el proyecto FAST [FAST 2001]), mientras que la validación de la Confiabilidad del sistema se realiza exclusivamente mediante técnicas de inyección (este es el caso del proyecto FIT [FIT 2002c]).…”