TOWARDS A MODELLING FRAMEWORK WITH TEMPORAL AND UNCERTAIN DATA FOR ADAPTIVE SYSTEMS

Doctoral thesis (2019)

Self-Adaptive Systems (SAS) optimise their behaviours or configurations at runtime in response to a modification of their environments or their behaviours. These systems therefore need a deep ... [more ▼]

Self-Adaptive Systems (SAS) optimise their behaviours or configurations at runtime in response to a modification of their environments or their behaviours. These systems therefore need a deep understanding of the ongoing situation which enables reasoning tasks for adaptation operations. Using the model-driven engineering (MDE) methodology, one can abstract this situation. However, information concerning the system is not always known with absolute confidence. Moreover, in such systems, the monitoring frequency may differ from the delay for reconfiguration actions to have measurable effects. These characteristics come with a global challenge for software engineers: how to represent uncertain knowledge that can be efficiently queried and to represent ongoing actions in order to improve adaptation processes? To tackle this challenge, this thesis defends the need for a unified modelling framework which includes, besides all traditional elements, temporal and uncertainty as first-class concepts. Therefore, a developer will be able to abstract information related to the adaptation process, the environment as well as the system itself. Towards this vision, we present two evaluated contributions: a temporal context model and a language for uncertain data. The temporal context model allows abstracting past, ongoing and future actions with their impacts and context. The language, named Ain’tea, integrates data uncertainty as a first-class citizen. [less ▲]

Enabling Temporal-Aware Contexts for Adaptative Distributed Systems

in SAC 2018: SAC 2018: Symposium on Applied Computing , April 9--13, 2018, Pau, France (2018)

Distributed adaptive systems are composed of federated entities offering remote inspection and reconfiguration abilities. This is often realized using a MAPE-K loop, which constantly evaluates system and ... [more ▼]

Distributed adaptive systems are composed of federated entities offering remote inspection and reconfiguration abilities. This is often realized using a MAPE-K loop, which constantly evaluates system and environmental parameters and derives corrective actions if necessary. The OpenStack Watcher project uses such a loop to implement resource optimization services for multi-tenant clouds. To ensure a timely reaction in the event of failures, the MAPE-K loop is executed with a high frequency. A major drawback of such reactivity is that many actions, e.g., the migration of containers in the cloud, take more time to be effective and their effects to be measurable than the MAPE-k loop execution frequency. Unfinished actions as well as their expected effects over time are not taken into consideration in MAPE-K loop processes, leading upcoming analysis phases potentially take sub-optimal actions. In this paper, we propose an extended context representation for MAPE-K loop that integrates the history of planned actions as well as their expected effects over time into the context representations. This information can then be used during the upcoming analysis and planning phases to compare measured and expected context metrics. We demonstrate on a cloud elasticity manager case study that such temporal action-aware context leads to improved reasoners while still be highly scalable. [less ▲]

Weaving Rules into Models@run.time for Embedded Smart Systems

in Mouline, Ludovic; Hartmann, Thomas; Fouquet, François (Eds.) et al Programming '17 Companion to the first International Conference on the Art, Science and Engineering of Programming (2017, April)

Smart systems are characterised by their ability to analyse measured data in live and to react to changes according to expert rules. Therefore, such systems exploit appropriate data models together with ... [more ▼]

Smart systems are characterised by their ability to analyse measured data in live and to react to changes according to expert rules. Therefore, such systems exploit appropriate data models together with actions, triggered by domain-related conditions. The challenge at hand is that smart systems usually need to process thousands of updates to detect which rules need to be triggered, often even on restricted hardware like a Raspberry Pi. Despite various approaches have been investigated to efficiently check conditions on data models, they either assume to fit into main memory or rely on high latency persistence storage systems that severely damage the reactivity of smart systems. To tackle this challenge, we propose a novel composition process, which weaves executable rules into a data model with lazy loading abilities. We quantitatively show, on a smart building case study, that our approach can handle, at low latency, big sets of rules on top of large-scale data models on restricted hardware. [less ▲]