Cloud computing has radically changed the way organisations operate their software by allowing them to achieve high availability of services at affordable cost. Containerized microservices is an enabling technology for this change, and advanced container orchestration platforms such as Kubernetes are used for service management. Despite the flourishing ecosystem of monitoring tools for such orchestration platforms, service management is still mainly a manual effort. The modeling of cloud computing systems is an essential step towards automatic management, but the modeling of cloud systems of such complexity remains challenging and, as yet, unaddressed. In fact modeling resource consumption will be a key to comparing the outcome of possible deployment scenarios. This paper considers how to derive resource models for cloud systems empirically. We do so based on models of deployed services in a formal modeling language with explicit CPU and memory resources; once the adherence to the real system is good enough, formal properties can be verified in the model. Targeting a likely microservices application, we present a model of Kubernetes developed in Real-Time ABS. We report on leveraging data collected empirically from small deployments to simulate the execution of higher intensity scenarios on larger deployments. We discuss the challenges and limitations that arise from this approach, and identify constraints under which we obtain satisfactory accuracy.

The advent of cloud computing radically changed the way organisations operate their applications and allows them to achieve high availability of services at affordable cost. Most cloud-computing platforms fostered Kubernetes for their container orchestration and service management. The scheduler is a key component of Kubernetes, as it is responsible for finding the placement of new service containers when they are deployed. The default scheduler is very fast, although often suboptimal. This can lead to inefficient placement of services, or more severely, inability to deploy.We present a custom Kubernetes scheduler, dubbed Boreas, which is designed to evaluate bursts of deployment requests concurrently. Boreas finds the optimal placements for service containers with their deployment constraints by utilising a configuration optimiser. Results show that Boreas is able to find placements where the default Kubernetes scheduler fails, wasting less computing resources, or proving that no feasible deployment solution is possible.

Loosely-coupled distributed systems organized as collections of so-called cloud-native microservices are able to adapt to traffic in very fine-grained and flexible ways. For this purpose, the cloud-native microservices exploit containerization and container management systems such as Kubernetes. This paper presents a formal model of resource consumption and scaling for containerized microservices deployed and managed by Kubernetes. Our aim is that the model, developed in Real-Time ABS, can be used as a framework to explore the behavior of deployed systems under various configurations at design time—before the systems are actually deployed. We further present initial results comparing the observed behavior of instances of our modeling framework to corresponding observations of real systems. These preliminary results suggest that the modeling framework can provide a satisfactory accuracy with respect to the behavior of distributed microservices managed by Kubernetes.

A Software Product Line (SPL) is a set of programs, called variants, which are generated from a common artifact base. Delta-Oriented Programming (DOP) is a flexible approach to implement SPLs. This short paper summarises the contributions published in [10]. A foundation for rigorous development of delta-oriented product lines of statecharts is provided by defining: a core language for statecharts, DOP on top of it, an analysis ensuring that a product line is well-formed (i.e., all variants can be generated and are well-formed statecharts). An implementation of the analysis has been applied to an industrial case study.

A Software Product Line (SPL) is a set of programs, called variants, which are generated from a common artifact base. Delta-Oriented Programming (DOP) is a flexible approach to implement SPLs. In this article, we provide a foundation for rigorous development of delta-oriented product lines of statecharts. We introduce a core language for statecharts, we define DOP on top of it, we present an analysis ensuring that a product line is well-formed (i.e., all variants can be generated and are well-formed statecharts), and we illustrate how an implementation of the analysis has been applied to an industrial case study. (C) 2018 Elsevier B.V. All rights reserved.