Power-Aware and Performance-Guaranteed Virtual Machine Placement in the Cloud

Zhao, Hui; Wang, Quan; Liu, Feng; Zhang, Weizhan; Zheng, Qinghua

doi:10.1109/tpds.2018.2794369

Cited by 117 publications

(56 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As Cloud computing is a promising technology, rapidly growing and appealing for industries, there is a large corpus of literature devoted to this topic. A large number of studies consider resource management in a single data center [6], [7], [43], [52], [53], [54], [55] while other studies focus on a more complex scenario where multiple data centers (and possibly providers) are involved [5], [56], [57], [58], [59]. For example, [5] focuses on placing cloud services over a distributed infrastructure by introducing affinity (from a QoS point of view) between services and geographic regions, data centers, up to single hosts.…”

Section: Related Workmentioning

confidence: 99%

“…For this to be achieved, advanced operations solutions are needed to provide support to performance prediction, monitoring of Service Level Agreements (SLAs), and adaptive configuration while satisfying requirements on cost-effectiveness, reliability, and security. Current solutions for enforcing SLAs mainly address single Clouds [6], [7], [8]; however, providing SLA guarantees with a…”

Section: Introductionmentioning

confidence: 99%

“…The third step (line 11) implements the receding horizon paradigm by modifying the application deployment using only the values calculated for the first time slot of the considered time window (to avoid greedy decisions and to exploit a better prediction for the steps not in the near future). Note that most of other literature approaches [6], [12], [30], [39] consider only a single step time instant in the future that can be obtained with n w = 1. Finally, (lines 12-15), the algorithm updates accordingly the system state, N j+t res,k,i , N j+t ond,k,i .…”

mentioning

confidence: 99%

See 2 more Smart Citations

A Hierarchical Receding Horizon Algorithm for QoS-Driven Control of Multi-IaaS Applications

Ardagna

Ciavotta

Lancellotti

et al. 2021

IEEE Trans. Cloud Comput.

View full text Add to dashboard Cite

Cloud Computing is emerging as a major trend in ICT industry. However, as with any new technology, new major challenges lie ahead, one of them concerning the resource provisioning. Indeed, modern Cloud applications deal with a dynamic context that requires a continuous adaptation process in order to meet satisfactory Quality of Service (QoS) but even the most titled Cloud platform provide just simple rule-based tools; the rudimentary autoscaling mechanisms that can be carried out may be unsuitable in many situations as they do not prevent SLA violations, but only react to them. In addition, these approaches are inherently static and cannot catch the dynamic behavior of the application and are unsuitable to manage multi-Cloud/data center deployments required for mission critical services. This situation calls for advanced solutions designed to provide Cloud resources in a predictive and dynamic way. This work presents capacity allocation algorithms, whose goal is to minimize the total execution cost while satisfying some constraints on the average response time of multi-Cloud based applications. This paper proposes a joint load balancing and receding horizon capacity allocation techniques, which can be employed to handle multiple classes of requests. An extensive evaluation of the proposed solution against an Oracle with perfect knowledge of the future and well-known heuristics proposed in the literature is provided. The analysis shows that our solution outperforms the heuristics producing results very close to the optimal ones, and reducing the number of QoS violations (in the worst case QoS constraints violation rate is 4.259% versus up to 17.245% of other approaches and can easily reduced by roughly a factor of four by exploiting the receding horizon approach). Furthermore, a sensitivity analysis over two different time scales indicates that finer grained time scales are more appropriate for spiky workloads. Analytical results are validated through simulation, which also analyzes the impact of Cloud environment random perturbations. Finally, experiments on a prototype environment demonstrate the effectiveness of the proposed approach under real workloads.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

A Hierarchical Receding Horizon Algorithm for QoS-Driven Control of Multi-IaaS Applications

Ardagna

Ciavotta

Lancellotti

et al. 2021

IEEE Trans. Cloud Comput.

View full text Add to dashboard Cite

show abstract

“…The coefficients α t,t can be estimated by monitoring tasks' performance as a function of their colocation and their sizes (a data center runs many instances of similar services). Previous works [20], [29], [36] and to some extent [37] show how to form a performance model as a function 1. To compute the cost of allocation, our model weights the loads by coefficients.…”

Section: Side-effects Of Colocating Tasks: a Modelmentioning

confidence: 99%

“…Analysis of effects of colocation. Studies showing performance degeneration when colocating data center tasks include [19], [21], [36], [37]. Xu et al [37] forms models of cloud tasks performance as a function of their resource consumption and colocation and proposes an optimization algorithm based on a metaheuristic.…”

Section: Related Workmentioning

confidence: 99%

Optimizing Egalitarian Performance when Colocating Tasks with Types for Cloud Data Center Resource Management

Pascual

Rządca

2019

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

In data centers, up to dozens of tasks are colocated on a single physical machine. Machines are used more efficiently, but the performance of the tasks deteriorates, as the colocated tasks compete for shared resources. Since the tasks are heterogeneous, the resulting performance dependencies are complex. In our previous work [26], [27] we proposed a new combinatorial optimization model that uses two parameters of a task -its size and its type -to characterize how a task influences the performance of other tasks allocated to the same machine.In this paper, we study the egalitarian optimization goal: the aim is to optimize the performance of the worst-off task. This problem generalizes the classic makespan minimization on multiple processors (P||C max ). We prove that polynomially-solvable variants of P||C max are NP-hard for this generalization, and that the problem is hard to approximate when the number of types is not constant. For a constant number of types, we propose a PTAS, a fast approximation algorithm, and a series of heuristics. We simulate the algorithms on instances derived from a trace of one of Google clusters. Compared with baseline algorithms solving P||C max , our proposed algorithms aware of the types of the jobs lead to significantly better tasks' performance. The notion of type enables us to extend standard combinatorial optimization methods to handle degradation of performance caused by colocation. Types add a layer of additional complexity. However, our results -approximation algorithms and good average-case performance -show that types can be handled efficiently.

show abstract

Energy‐aware andSLA‐guaranteed optimal virtual machine swap and migrate system in cloud‐Internet of Things

Karthikeyan¹,

Balamurugan²

2021

Concurrency and Computation

View full text Add to dashboard Cite

An emerging cloud‐Internet of Things (IoT) is a novel paradigm that brings potential benefits over a variety of applications. The pervasive use of IoT devices often struggle to meet resource requirements in cloud environment. The abundant wastage or inappropriate usage of resources leads to consume larger amount of energy and delayed response. Optimization algorithms are more popular for optimal selection, the algorithms as ant colony optimization, particle swarm optimization, genetic algorithm were majorly concentrated for the purpose of load balancing. However workload is balanced in many previous research works, however it failed to mitigate SLA violations and limitations of energy consumption. In this paper, we address both energy‐aware load balancing and satisfaction of SLA constraints. First, IoT devices submit tasks which are segregated into queues using policy‐based SLA by taking in account of SLA constraints. Second, the tasks are allocated in accordance to dynamic threshold predicting fuzzy followed by Analytical Hierarchical Process. Third, the workload is employed for energy minimization which decides whether to perform swap or migrate virtual machines (VMs). Swapping between VMs is held by novel map and consolidates processes. Then combination of fruit fly and bird swarm hybrid optimization algorithm is enabled to select an optimal VM for workload balancing. Also, the idle physical machines (PMs) are supposed to be in OFF state for diminishing unnecessary energy consumption. The outcomes of this cloud‐IoT system is experimentally evaluated and compared in terms of energy consumption, number of migrations, resource utilization, and execution time.

show abstract

Power-Aware and Performance-Guaranteed Virtual Machine Placement in the Cloud

Cited by 117 publications

References 47 publications

A Hierarchical Receding Horizon Algorithm for QoS-Driven Control of Multi-IaaS Applications

A Hierarchical Receding Horizon Algorithm for QoS-Driven Control of Multi-IaaS Applications

Optimizing Egalitarian Performance when Colocating Tasks with Types for Cloud Data Center Resource Management

Energy‐aware andSLA‐guaranteed optimal virtual machine swap and migrate system in cloud‐Internet of Things

Contact Info

Product

Resources

About