Robust Dynamic CPU Resource Provisioning in Virtualized Servers

Makridis, Evagoras; Deliparaschos, Kyriakos M.; Kalyvianaki, Evangelia; Zolotas, Argyrios; Charalambous, Themistoklis

doi:10.1109/tsc.2020.2966972

Cited by 16 publications

(8 citation statements)

References 46 publications

(62 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In general, the cost of workloads assigned by ( 41) is always less than (or equal to) the cost associated with the balanced model. We add some penalty functions to keep the servers at an operating point away from the capacity, in fact, below 70 − 80% of their capacity (due to the uncertainty of the processing times) since the Mean Response Time of the servers grows (exponentially) at some point [54]. As a ruleof-thumb, we address this concern by box constraints on the load-to-capacity ratios as 0 ≤ z i ≤ 0.75 (with z i = wi+ρi 23) gives the max step rate η = 4.85.…”

Section: Simulations a Application: Optimizing The Cpu Utilizationmentioning

confidence: 99%

Distributed Anytime-Feasible Resource Allocation Subject to Heterogeneous Time-Varying Delays

Doostmohammadian

Aghasi

Rikos

et al. 2022

IEEE Open J. Control. Syst.

Self Cite

View full text Add to dashboard Cite

This paper considers distributed allocation strategies, formulated as a distributed sumpreserving (fixed-sum) allocation of resources over a multi-agent network in the presence of heterogeneous arbitrary time-varying delays. We propose a double time-scale scenario for unknown delays and a faster single time-scale scenario for known delays. Further, the links among the nodes are considered subject to certain nonlinearities (e.g, quantization and saturation/clipping). We discuss different models for nonlinearities and how they may affect the convergence, sum-preserving feasibility constraint, and solution optimality over general weight-balanced uniformly strongly connected networks and, further, time-delayed undirected networks. Our proposed scheme works in a variety of applications with general non-quadratic strongly-convex smooth objective functions. The non-quadratic part, for example, can be due to additive convex penalty or barrier functions to address the local box constraints. The network can change over time, is not necessarily connected at all times, but is only assumed to be uniformly-connected. The novelty of this work is to address all-time feasible Laplacian gradient solutions in presence of nonlinearities, switching digraph topology (not necessarily all-time connected), and heterogeneous time-varying delays.

show abstract

Section: Simulations a Application: Optimizing The Cpu Utilizationmentioning

confidence: 99%

Distributed Anytime-Feasible Resource Allocation Subject to Heterogeneous Time-Varying Delays

Doostmohammadian

Aghasi

Rikos

et al. 2022

IEEE Open J. Control. Syst.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Makridis et al [44] present robust dynamic resource allocation mechanisms to allocate application resources meeting Service Level Objectives (SLOs) agreed between Cloud providers and customers. The controllers are self-adaptive, with process noise variances and covariances calculated using previous measurements within a time window.…”

Section: Adaptive Resource Allocationmentioning

confidence: 99%

Efficient Resource Management on Container as a Service

Park¹

2022

Preprint

View full text Add to dashboard Cite

Docker has been widely adopted as a platform solution for microservice. As the popularity of microservice increases, the importance of fine-tuning the efficiency of resource management in the Docker platform also increases. While Docker’s out-of-box resource management solution provides some generic management capability, more work is required to improve resource utilization and enforce Service Level Agreement (SLA) for critical services. In this research, an efficient Docker resource management scheme, called Adaptive SLA Enforcement, is designed and implemented. For the sake of comparison, we also study and implement three simpler schemes: 1) Fixed Number of Containers, 2) Dynamic Resource Management without SLA Enforcement, 3) Strict SLA Enforcement. We found that the Adaptive SLA Enforcement scheme can deliver efficient resource management with SLA enforcement, thus successfully addressing the deficiencies of the other three schemes.

show abstract

“…These conclusions are drawn for the process, fuzzy controller, optimization problem and dynamic regime considered in this paper. Other processes, controllers, optimization problems and dynamic regimes are expected to lead to different conclusions; challenging processes in this regard are those considered in [57][58][59][60][61][62][63][64].…”

Section: Validation and Comparisonmentioning

confidence: 99%

Slime Mould Algorithm-Based Tuning of Cost-Effective Fuzzy Controllers for Servo Systems

Precup¹,

David²,

Roman³

et al. 2021

IJCIS

107

View full text Add to dashboard Cite

This paper suggests five new contributions with respect to the state-of-the-art. First, the optimal tuning of cost-effective fuzzy controllers represented by Takagi-Sugeno-Kang proportional-integral fuzzy controllers (TSK PI-FCs) is carried out using a fresh metaheuristic algorithm, namely the Slime Mould Algorithm (SMA), and a fuzzy controller tuning approach is offered. Second, a relatively easily understandable formulation of SMA is offered. Third, a real-world application of SMA is given, focusing on the optimal tuning of TSK PI-FCs for nonlinear servo systems in terms of optimization problems that target the minimization of discrete-time cost functions defined as the sum of time multiplied by squared control error. Fourth, using the concept of improving the performance of metaheuristic algorithms with information feedback models, proposed by Wang and Tan, Improving metaheuristic algorithms with information feedback models, IEEE Trans. Cybern. 49 (2019), 542-555, Gu and Wang, Improving NSGA-III algorithms with information feedback models for large-scale many-objective optimization, Fut. Gen. Comput. Syst. 107 (2020), 49-69, and Zhang et al., Enhancing MOEA/D with information feedback models for large-scale many-objective optimization, Inf. Sci. 522 (2020), 1-16, new metaheuristic algorithms are introduced in terms of inserting the model F1 in SMA and other representative algorithms, namely Gravitational Search Algorithm (GSA), Charged System Search (CSS), Grey Wolf Optimizer (GWO) and Whale Optimization Algorithm (WOA). Fifth, the real-time validation of the cost-effective fuzzy controllers and their tuning approach is performed in the framework of angular position control of laboratory servo system. The comparison with other metaheuristic algorithms that solve the same optimization problem for optimal parameter tuning of cost-effective fuzzy controllers suggestively highlights the superiority of SMA. Experimental results are included.

show abstract

Robust Dynamic CPU Resource Provisioning in Virtualized Servers

Cited by 16 publications

References 46 publications

Distributed Anytime-Feasible Resource Allocation Subject to Heterogeneous Time-Varying Delays

Distributed Anytime-Feasible Resource Allocation Subject to Heterogeneous Time-Varying Delays

Efficient Resource Management on Container as a Service

Slime Mould Algorithm-Based Tuning of Cost-Effective Fuzzy Controllers for Servo Systems

Contact Info

Product

Resources

About