Resource-utilization-aware energy efficient server consolidation algorithm for green computing in IIOT

https://doi.org/10.1016/j.jnca.2017.07.011Get rights and content

Highlights

  • The proposed RUAEE in this paper entails four parts: disposition of overloaded hosts, adjustment of unbalance hosts, selection of underloaded hosts and VM placement to improve hosts’ resource utilization and reduce the number of SLA violations.

  • At the beginning of server consolidation, overloaded hosts are disposed for improving the Quality of Service.

  • The resource utilization description model is proposed for identifying unbalance hosts and unbalance hosts are selected for workload adjustment.

  • And then, the underloaded host selection module detects low-utilized hosts by Select Factor (SF) value and migrates all of their running VMs out for switching them off to achieve energy saving.

  • At last, migrating VMs are elaborately placed on suitable PMs to improve destination hosts’ resource utilization.

Abstract

Cloud computing has become an indispensable infrastructure that provides multi-granularity services to support large applications in the Industrial Internet of Things (IIOT), Cloud data centers have been built or extensively enlarged to cope with the growing computation and storage requirements of IIOT. The energy consumption of cloud data centers is dramatically increasing, which has created a lot of problems with greenhouse gas emissions and service costs. Server consolidation is a popular approach to reduce cloud data centers’ energy consumption by minimizing the number of active physical machines. Most of the extant research has focused on server reduction in the consolidation process, but unbalanced resource utilization among different physical machines can cause the waste of physical resources. This paper proposes a resource-utilization-aware energy efficient server consolidation algorithm (RUAEE) that can be used to improve resource utilization while reducing the number of virtual machine live migrations. Experimental results show that RUAEE can reduce the energy consumption and service-level agreement (SLA) violation in cloud data center.

Introduction

The cloud computing paradigm (Zhang et al., 2016) has significantly promoted the rapid development of the Industrial Internet of Things (IIOT) (Jeschke et al., 2017) and currently provides flexible, extensible services to its subscribers through a pay-as-you-go model. Cloud users are free from the regular configuration of hardware and software systems, and have no geographic restriction in accessing cloud services; they can freely access Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS) over the internet (Jeyarani et al., 2012). The computing and storage provided by cloud computing allow for many useful applications such as decision-making systems and IIOT intelligent optimization, which serve control systems and the manufacturing industry. The rapid development of IIOT has also supported further advancements in cloud computing.

The number of cloud data centers worldwide has dramatically increased in recent years. In 2010, the energy consumed by cloud data centers almost accounts for 1.1–1.5% of the world's (Koomey, 2011). The huge power consumption not only increases greenhouse gas emissions contributing to global warming, but also indirectly increases the costs passed down to cloud users. Further, according to statement in Barroso and Holzle (2007) and NY Times (2014), the resource utilization in cloud data centers is remarkably low, usually between 10% and 50% of its capacity. Improving the resource utilization of physical nodes and minimizing the number of active servers are worthwhile approaches to reducing the energy consumption of cloud data centers.

Virtualization technology is a fine-grained solution to reallocate physical resources between different physical nodes in cloud data center. It provides identical visualizations of diverse physical resources by creating normalized Virtual Machines (VMs) on Physical Machines (PMs). Applications can be submitted to different VMs for concurrently running on the same PM. Moreover, live migration can be used to aggregate VMs into few PMs to reduce the quantity of active servers; this is concisely referred to as “server consolidation”.

Server consolidation in cloud data center is usually treated as a bin-packing problem, in which physical nodes are considered as bins and VMs as items. Classic bin-packing algorithms are not particularly applicable to server consolidation because the workload in the cloud continually and dynamically changes,causing continual and dynamic VM migrations. Live migrations can degrade service performance and violate the Service Level Agreement (SLA), which is a kind of metrics for the Quality of Service (QoS). Hence, it is necessary to reduce the amount of VM migrations for ensuring service performance when designing heuristic server consolidation algorithm.

Resource fragmentation is another notable problem cased by server consolidation. Different VMs require different manner of physical resources, (e.g., CPU, memory), a PM's resource utilization can be unbalanced and it becomes unavailable for in-coming VMs if even just one dimensional physical resource is exhausted Li et al., 2013, Wei et al.,. The unutilized physical resources in the unavailable PM, i..e, “resource fragments”, are just wasted. As shown in Fig. 1(a), if a given PM's memory utilization is much higher than its CPU utilization, then the PM cannot readily accommodate other VMs and nearly half of its CPU resources are wasted. Fig. 1(b) depicts resource fragment in the opposite circumstances, where almost half of the memory is wasted. It is crucial to reduce the amount of resource fragments as much as possible to improve the utilization of physical nodes through server consolidation.

In this paper, we take the above mentioned problems into consideration and propose a resource-utilization-aware energy efficient server consolidation algorithm (RUAEE) to improve physical nodes’ resource utilization and reduce energy consumption in cloud data centers. The main contributions of this paper can be summarized as follows.

  • An efficient approach is proposed for selecting underutilized hosts for consolidation in order to reduce the amount of VM live migrations.

  • A descriptive resource utilization model is proposed to guide the adjustment of unbalanced hosts and VM placements to improve resource utilization and minimize the active number of PMs.

  • Real workload traces from the Google cluster are used to run a series of simulations to validate the proposed algorithm.

The remainder of this paper is organized as follows. Section 2 discusses related work, and the problem statement is presented in Section 3. The proposed algorithm is discussed in detail in 4 Resource utilization description model, 5 Proposed algorithm. The experiments and results are discussed in 6 Experiments and results, 7 Conclusion summarizes our conclusions and suggests future research directions.

Section snippets

Related work

Green computing has been a popular research subject for several years. Reducing energy consumption in the cloud infrastructure is the focus of many studies on this subject. Nathuji and Schwan (2007) conducted one of the first studies on power management in virtualized data centers. The management system is composed of both a local and global manager; The global manager is tasked with coordinating the guest OSs power management strategies, while the local manager gathers the host's performance

Problem statement

The server consolidation problem conducted by VM live migration is usually treated as a bin-packing problem. There are important differences to consider, however, when designing relevant solutions. Bin packing algorithms are generally built under the assumption that all bins are empty, but server consolidation algorithms must be tailored to cloud data centers where VMs have been created on physical nodes. Minimizing the number of active physical nodes for energy-saving purposes in cloud data

Resource utilization description model

Because physical nodes can be unavailable for migrating VMs when even one dimensional resource does not satisfy the VMs resource requirement, these unavailable hosts’ residual physical resources are simply wasted during its life time. The unbalanced physical nodes also have a negative effect on energy consumption in the cloud data center overall. In order to describe these resource fragments and improve the host's resource utilization, we propose a resource utilization description model that

Proposed algorithm

We designed RUAEE for green computing in IIOT. It can reduce energy consumption and the number of VM live migrations by periodically performing consolidation. The proposed solution can be divided into four parts: disposition of overloaded hosts, adjustment of unbalanced hosts, selection of underloaded hosts, and VM placement. At the beginning of every consolidation interval, the module of overloaded disposition begins to manage overloaded hosts and classify active servers for other modules.

Experimental setup

As the target system is in IaaS provides cloud services via large scale virtualized data center infrastructures, the proposed algorithm and performance evaluation should ideally be tested in a similar scenario. Actually implementing the consolidation algorithms on a real infrastructure for large-scale experiments is extremely difficult, however. We choose the CloudSim toolkit (Calheiros et al., 2011) as a simulation platform to evaluate the performance of the proposed algorithm to ensure

Conclusion

Server consolidation is an effective approach to improving resource utilization and reducing power consumption in cloud data centers. It can be regarded as a type of bin-packing problem, where VMs represent items and nodes represent bins; active PMs denotes bins which should be minimized. The consolidation algorithm cannot start at an empty state, so it is important to take current state into consideration in attempting to reduce the number of VM live migrations.

In this paper, we took both the

Acknowledgements

The work is supported by “the National Natural Science Foundation of China under Grant nos. 61572172 and 61602137” and supported by “the Fundamental Research Funds for the Central Universities, No. 2016B10714 and supported by “Changzhou Sciences and Technology Program, Nos. CE20165023 and CE20160014” and “Six Talent peaks project in Jiangsu Province, No. XYDXXJS-007”.

References (30)

  • D. Bein et al.

    Cloud storage and online bin packing

    Intell. Distrib. Comput. V

    (2011)
  • Beloglazov, A., Buyya, R., 2010. Energy efficient allocation of virtual machines in cloud data centers, cluster, cloud...
  • A. Beloglazov et al.

    Optimal online deterministic algorithms and adaptive heuristics for energy and performance efficient dynamic consolidation of virtual machines in cloud data centers

    Concurr. Comput.: Pract. Exp.

    (2012)
  • K. Bilal et al.

    Trends and challenges in cloud datacenters

    Growth

    (2014)
  • R.N. Calheiros et al.

    CloudSim: a toolkit for modeling and simulation of cloud computing environments and evaluation of resource provisioning algorithms

    Softw.: Pract. Exp.

    (2011)
  • View full text