Beyond the rainbow: self-adaptive failure avoidance in configurable systems

Abstract

Self-adaptive software systems monitor their state and then adapt when certain conditions are met, guided by a global utility function. In prior work we developed algorithms and conducted a post-hoc analysis demonstrating the possibility of adapting to software failures by judiciously changing configurations. In this paper we present the REFRACT framework that realizes this idea in practice by building on the self-adaptive Rainbow architecture. REFRACT extends Rainbow with new components and algorithms targeting failure avoidance. We use REFRACT in a case study running four independently executing Firefox clients with 36 passing test cases and 7 seeded faults. The study show that workarounds for all but one of the seeded faults are found and the one that is not found never fails -it is guarded from failing by a related workaround. Moreover, REFRACT finds workarounds for eight configuration-related unseeded failures from tests that were expected to pass (and did under the default configuration). Finally, the data show that when a failure and its workaround are found, configuration guards prevent the failure from appearing again. In a simulation lasting 24 hours we see over 150 guard activations and no failures with workarounds remaining beyond 16 hours.

MORPH: a reference architecture for configuration and behaviour self-adaptation

“…Note that strategies do not have to be necessarily temporal sequencing of actions or commands. For instance, in [55] reconfiguration strategies used are one-step component parameter changes.…”

Runtime Failure Prevention and Reaction

“…The general intuition is that these techniques may try to replay a failing execution many times until finding a change that might be operated on the monitored program to prevent the failure without breaking the other functionalities of the system. The change might be either on the configuration [90] or in the code [46] of the monitored program. Of course, a suitable environment is needed to replay an execution many times regardless of the side-effects that might be introduced by a failing execution.…”

“…REFRACT implements this strategy using a feature model as representation of the configuration space [90]. In practice, when a failure is observed, REFRACT replays the failed execution in a separate environment and samples the configuration space described by the feature model according to three possible strategies: n-hops sampling, random sampling, and covering array sampling.…”

An Extended Description of MORPH: A Reference Architecture for Configuration and Behaviour Self-Adaptation

“…Note that strategies do not have to be necessarily temporal sequencing of actions or commands. For instance, in [20] reconfiguration strategies used are one-step component parameter changes.…”