Daniel Moldovan

SYBL: An Extensible Language for Controlling Elasticity in Cloud Applications

Elasticity in cloud computing is a complex problem, regarding not only resource elasticity but also quality and cost elasticity, and most importantly, the relations among the three. Therefore, existing support for controlling elasticity in complex applications, focusing solely on resource scaling, is not adequate. In this paper we present SYBL - a novel language for controlling elasticity in cloud applications - and its runtime system. SYBL allows specifying in detail elasticity monitoring, constraints, and strategies at different levels of cloud applications, including the whole application, application component, and within application component code. Based on simple SYBL elasticity directives, our runtime system will perform complex elasticity controls for the client, by leveraging underlying cloud monitoring and resource management APIs. We also present a prototype implementation and experiments illustrating how SYBL can be used in real-world scenarios.

Multi-level Elasticity Control of Cloud Services

Fine-grained elasticity control of cloud services has to deal with multiple elasticity perspectives quality, cost, and resources. We propose a cloud services elasticity control mechanism that considers the service structure for controlling the cloud service elasticity at multiple levels, by firstly defining an abstract composition model for cloud services and enabling multi-level elasticity control. Secondly, we define mechanisms for solving conflicting elasticity requirements and generating action plans for elasticity control. Using the defined concepts and mechanisms we develop a runtime system supporting multiple levels of elasticity control and validate the resulted prototype through experiments.

MELA: Monitoring and Analyzing Elasticity of Cloud Services

Cloud computing has enabled a wide array of applications to be exposed as elastic cloud services. While the number of such services has rapidly increased, there is a lack of techniques for supporting cross-layered multi-level monitoring and analysis of elastic service behavior. In this paper we introduce novel concepts, namely elasticity space and elasticity pathway, for understanding elasticity of cloud services, and techniques for monitoring and evaluating them. We present MELA, a customizable framework, which enables service providers and developers to analyze cross-layered, multi-level elasticity of cloud services, from the whole cloud service to service units, based on service structure dependencies. Besides support for real-time elasticity analysis of cloud service behavior, MELA provides several customizable features for extracting functions and patterns that characterize that behavior. To illustrate the usefulness of MELA, we conduct several experiments with a realistic data-as-a-service in an M2M cloud platform.

ADVISE – A Framework for Evaluating Cloud Service Elasticity Behavior

Complex cloud services rely on different elasticity control processes to deal with dynamic requirement changes and workloads. However, enforcing an elasticity control process to a cloud service does not always lead to an optimal gain in terms of quality or cost, due to the complexity of service structures, deployment strategies, and underlying infrastructure dynamics. Therefore, being able, a priori, to estimate and evaluate the relation between cloud service elasticity behavior and elasticity control processes is crucial for runtime choices of appropriate elasticity control processes. In this paper we present ADVISE, a framework for estimating and evaluating cloud service elasticity behavior. ADVISE gathers service structure, deployment, service runtime, control processes, and cloud infrastructure information. Based on this information, ADVISE utilizes clustering techniques to identify cloud elasticity behavior produced by elasticity control. Our experiments show that ADVISE can estimate the expected elasticity behavior, in time, for different cloud services thus being a useful tool to elasticity controllers for improving the quality of runtime elasticity control decisions.

Energy Aware Dynamic Resource Consolidation Algorithm for Virtualized Service Centers Based on Reinforcement Learning

In this paper we propose an energy aware dynamic consolidation algorithm for virtualized service centers based on reinforcement learning. The energy awareness is enacted by using the Energy Aware Context Model (EACM) to programmatically represent the current service center context situation by means of ontologies. We have defined the EACM model entropy metric for evaluating the service center greenness level. If the entropy value is above a predefined threshold, the service center is not in a green state. As a consequence, consolidation or dynamic power management actions are selected by means of reinforcement learning and executed to bring back the service center in an energy efficient state. The results are promising showing that the proposed energy aware consolidation algorithm decreases the energy consumption with about 26% from the total energy consumption of a service center.

Cloud SLA negotiation for energy saving — A particle swarm optimization approach

The necessity of balancing the obtained performance with the energy consumed is an emerging ambition for cloud computing research. Performance in cloud computing is defined through Service Level Agreement contracts between the cloud provider and cloud customer, being a projection of the customers perspective on the service offered by the cloud provider. Although more and more research efforts go into standardizing Service Level Agreement in cloud systems, the area is still at its early ages. This paper proposes a Service Level Agreement negotiation protocol based on particle swarm optimization techniques, for obtaining a balance between the energy consumed and performance offered in the cloud. The two parties of the defined negotiation protocol are the performance-oriented cloud customer and the energy-oriented cloud provider. The agreement resulted from the negotiation process satisfies the two major negotiation properties we aim for: closeness to Pareto optimality and high social welfare.

CoMoT -- A Platform-as-a-Service for Elasticity in the Cloud

Platform-as-a-Service (PaaS) should support the design, deployment, execution, test and monitoring of native elastic systems constructed from elastic service units based on multi-dimensional elasticity requirements. In this paper, we discuss fundamental building blocks for enabling multi-dimensional elasticity programming of software-defined elastic systems. We describe CoMoT, a novel PaaS for elasticity in the cloud that is developed based on these fundamental building blocks.

On estimating actuation delays in elastic computing systems

Elastic controllers autonomically adjust the allocation of resources in cloud computing systems. Usually such controllers assume that control actions will take immediate effect. In clouds, however, actuation times may be long, and the controllers can hardly guarantee acceptable levels of service if they neglect these actuation delays. Therefore, the ability to correctly estimate the time that control actions take effect on the systems is crucial. However, detecting actuation delays in elastic computing systems is challenging because cloud systems provide only inaccurate and incomplete data about reconfigurations timing. In this paper, we tackle the problem of estimating the delay of control actions in elastic systems. We identify recurring types of changes in the monitored metrics and requirements to properly carry out the estimation. Based on that, we develop a novel framework for the actuation delays estimation that utilizes change point detection techniques. We conduct several experiments with real-world systems to illustrate the feasibility and applicability of our framework.

On Controlling Cloud Services Elasticity in Heterogeneous Clouds

Various complex cloud services have to be deployed in multiple heterogeneous clouds, due to the service requirements for particular functionalities from specific clouds. In order to control these cloud services, we need to monitor and control the various units deployed across multiple clouds, dealing with cloud-specific protocols to support an end-to-end cloud service perspective. In this paper we present an approach for multi-cloud control, which evaluates relationships among different units deployed across heterogeneous clouds, and generates action plans necessary for controlling service elasticity. We show experiments of the end-to-end control and sensitivity analysis for a service deployed across two different types of clouds.

An energy aware context model for green IT service centers

In this paper we propose the development of an Energy Aware Context Model for representing the service centre energy/performance related data in a uniform and machine interpretable manner. The model is instantiated at run-time with the service center energy/performance data collected by monitoring tools. Energy awareness is achieved by using reasoning processes on the model instance ontology representation to determine if the service center Green and Key Performance Indicators (GPIs/KPIs) are fulfilled in the current context. If the predefined GPIs/KPIs are not fulfilled, the model is used as primary resource to generate run-time adaptation plans that should be executed to increase the service centers greenness level.