The fast increasing presence of Internet-of-Things and fog computing resources exposes new challenges due to heterogeneity and non-negligible network delays among resources as well as the dynamism of operating conditions. Such a variable computing environment leads the applications to adopt an elastic and decentralized execution. To simplify the application deployment and run-time management, containers are widely used nowadays. The deployment of a container-based application over a geo-distributed computing infrastructure is a key task that has a significant impact on the application non-functional requirements (e.g., performance, security, cost). In this survey, we first develop a taxonomy based on the goals, the scope, the actions, and the methodologies considered to adapt at run-time the application deployment. Then, we use it to classify some of the existing research results. Finally, we identify some open challenges that arise for the application deployment in the fog. In literature, we can find many different approaches for adapting the containers deployment, each tailored for optimizing a specific objective, such as the application response time, its deployment cost, or the efficient utilization of the available computing resources. However, although several solutions for deploying containers exist, those explicitly considering the distinctive features of fog computing are at the early stages: indeed, existing solutions scale containers without considering their placement, or do not consider the heterogeneity, the geographic distribution, and mobility of fog resources.
CITATION STYLE
Cardellini, V., Lo Presti, F., Nardelli, M., & Rossi, F. (2020). Self-adaptive Container Deployment in the Fog: A Survey. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 12041 LNCS, pp. 77–102). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-58628-7_6
Mendeley helps you to discover research relevant for your work.