Dynamic Resource Reallocation between Deployment Components

Abstract

Abstract. Today's software systems are becoming increasingly configurable and designed for deployment on a plethora of architectures, ranging from sequential machines via multicore and distributed architectures to the cloud. Examples of such systems are found in, e.g., software product lines, service-oriented computing, information systems, embedded systems, operating systems, and telephony. To model and analyze systems without a fixed architecture, the models need to naturally capture and range over relevant deployment scenarios. For this purpose, it is interesting to lift aspects of low-level deployment concerns to the abstraction level of the modeling language. In this paper, the objectoriented modeling language Creol is extended with a notion of dynamic deployment components with parametric processing resources, such that processor resources may be explicitly reallocated. The approach is compositional in the sense that functional models and reallocation strategies are both expressed in Creol, and functional models can be run alone or in combination with different reallocation strategies. The formal semantics of deployment components is given in rewriting logic, extending the semantics of Creol, and executes on Maude, which allows simulations and test suites to be applied to models which vary in their available resources as well as in their resource reallocation strategies.

Validating Timed Models of Deployment Components with Parametric Concurrency

“…It is a timed extension of ABS [17], which builds on the Creol language [18] (used in our previous work [19,20]) but provides a fully implemented functional sublanguage and support for software product line modeling [7]. Characteristic features of Timed ABS are that: (1) it allows abstracting from implementation details while remaining executable; i.e., a functional sublanguage over abstract data types is used to specify internal, sequential computations [17]; (2) it provides flexible concurrency and synchronization mechanisms by means of asynchronous method calls, release points in method definitions, and cooperative scheduling of method activations [9,18]; (3) it supports user-provided deadlines to method calls to express local QoS requirements [10]; and (4) it features deployment components with parametric resources to model deployment variability [2,19,20]. Compared to previous work on deployment components, we here extend the syntax and semantics of Timed ABS with user-defined annotations to express general cost-models for resource usage.…”

“…This compositionality allows concurrent objects to be naturally distributed on different locations, because only the local state of a concurrent object is needed to execute its methods. In previous work [2,19,20], the authors have introduced deployment components as a modeling concept which captures restricted resources shared between a group of concurrent objects, and shown how components with parametric resources may be used to capture a model's behavior for different assumptions about the available resources.…”

“…In recent work, we have extended ABS with time and with a notion of deployment component in order to abstractly capture resource restrictions related to deployment decisions at the modeling level. This allows us to observe, by means of simulations, the performance of a system model ranging over the amount of resources assigned to the deployment components, with respect to execution capacity [19,20] and with respect to memory [2]. This way, the modeler gains insight into the resource requirements of a component, in order to provide a minimum response time for given client behaviors.…”

Modeling Application-Level Management of Virtualized Resources in ABS

“…This explicit representation of deployment architecture allows application-level response time and load balancing to be expressed in the system models in a very natural and flexible way, relative to the resources allocated to an application. This basic model of deployment architecture may be further extended by adding support for resource reallocation [23] or object mobility [25], allowing resources or objects to be moved from one deployment component to another.…”

“…The formal syntax and semantics of ABS is discussed in [21] and the formalization of the integration of deployment components with ABS in [23][24][25]. In this paper we show how deployment components in ABS may be used to model virtualized systems, by developing an example inspired by cloud computing [8].…”