Service Level Agreement Requirements Research Articles

Prediction-based Resource Provision and Virtual Machines Placement in Cloud Data Center

With the rapid development of virtualization, cloud servers provide the power and flexibility that single servers struggle to provide. However, low utilization of physical machines and high energy consumption are the main concerns for cloud service providers. In this paper, we propose a predict-and-place framework to decrease the number of active physical machines in cloud centers. We analyze the historical VM (virtual machine) request records to predict the VM demands in the next several days and then use an offline VM placement strategy to keep the number of active servers as less as possible while satisfying the SLA (Service-Level Agreement) requirements. In the prediction process, we enhanced the classical Holt’s linear prediction method on account of inevitable outliers to increase the prediction accuracy. We conduct experiments to compare the prediction accuracy of Perceptron, classical Holt-Winters, and our refined, robust Holt-Winters method. The experiment based on real data shows the simple perceptron has an inferior result. Our improved prediction method reduces the MAPE (Mean Absolute Percentage Error) by about 50% compared with the classical one, and the predict-and-place framework decreases the average number of active physical machines effectively in the cloud data center.

Resource needs prediction in virtualized systems: Generic proactive and self-adaptive solution

Resource management of virtualized systems in cloud data centers is a critical and challenging task due to the fluctuating workloads and complex applications in such environments. Over-provisioning is a common practice to meet service level agreement requirements, but this leads to under-utilization of resources and energy waste. Thus, provisioning virtualized systems with resources according to their workload demands is essential. Existing solutions fail to provide a complete solution in this regard, as some of them lack proactivity and dynamism in estimating resources, while others are environment- or application-specific, which limits their accuracy in the case of bursty workloads. Effective resource management requires dynamic and accurate prediction. This work presents a novel prediction algorithm, which (1) is generic, and can thus be applied to any virtualized system, (2) is able to provide proactive estimation of resource requirements through machine learning techniques, and (3) is capable of real-time adaptation with padding and prediction adjustments based on prediction error probabilities in order to reduce under- and over-provisioning of resources. In several virtualized systems, and under different workload profiles, the experimental results show that our proposition is able to reduce under-estimation by an average of 86% over non-adjusted prediction, and to decrease over-estimation by an average of 67% versus threshold-based provisioning.

A Complete LTE Mathematical Framework for the Network Slice Planning of the EPC

5G is the next telecommunications standards that will enable the sharing of physical infrastructures to provision ultra short-latency applications, mobile broadband services, Internet of Things, etc. Network slicing is the virtualization technique that is expected to achieve that, as it can allow logical networks to run on top of a common physical infrastructure and ensure service level agreement requirements for different services and applications. In this vein, our paper proposes a novel and complete solution for planning network slices of the LTE EPC, tailored for the enhanced Mobile BroadBand use case. The solution defines a framework which consists of: i) an abstraction of the LTE workload generation process, ii) a compound traffic model, iii) performance models of the whole LTE network, and iv) an algorithm to jointly perform the resource dimensioning and network embedding. Our results show that the aggregated signaling generation is a Poisson process and the data traffic exhibits self-similarity and long-range-dependence features. The proposed performance models for the LTE network rely on these results. We formulate the joint optimization problem of resources dimensioning and embedding of a virtualized EPC and propose a heuristic to solve it. By using simulation tools, we validate the proper operation of our solution.

IoT Public Fog Nodes Reputation System: A Decentralized Solution Using Ethereum Blockchain

Public fog nodes extend cloud services for the Internet of Things (IoT) clients and smart devices to provide additional computation capabilities, storage space, and reduce latency and response time. The openness and pervasiveness of public fog nodes leads to the requirement of using trust models to ensure reliability, security, privacy, and meet the service-level agreements (SLAs). Conventional trust models for public fog nodes are centrally configured, deployed, and maintained considering security, privacy, and SLA requirements. However, these trust models enforce centralized governance policies across the system which leads towards the single-point-of-failure and single-point-of-compromise over IoT devices’ and users’ personal data. This paper proposes a decentralized trust model in order to maintain the reputation of publicly available fog nodes. The reputation is maintained considering users’ opinions about their past interactions with the public fog nodes. The proposed trust model is designed using public Ethereum blockchain and smart contract technologies in order to enable decentralized trustworthy service provisioning between IoT devices and public fog nodes. The proposed approach is tested and evaluated in terms of security, performance, and cost. The results show that using blockchain for decentralized reputation management could become more advantageous when compared to the existing centralized trust models.

Cost‐minimizing online algorithm for internet green data centers on multi‐source energy

SummaryHuge energy consumption of large‐scale cloud data centers damages environments with excessive carbon emission. More and more data center operators are seeking to reduce carbon footprint via various types of renewable energy. However, the intermittent availability of renewable energy sources makes it quite challenging to cooperate with dynamically arriving workload. Meanwhile, the different natures (eg, price and carbon emission) of multiple energy sources also bring more challenges to achieve an optimal trade‐off among carbon emission, power cost, and service level agreement (SLA). In this paper, we study the problem of reducing the long‐term energy cost for geo‐distributed cloud centers, where multiple sources of renewable energy are considered and SLA requirement and carbon budget are satisfied. To tackle the randomness of workload arrival, varying electricity price, and intermittent supply of renewable energy, we first formulate the cost minimization problem as a constraint stochastic optimization problem. Second, based on Lyapunov optimization technique, we propose an online control algorithm to solve it and provide the rigorous theory analysis to demonstrate its performance. By converting the long‐term optimization problem to a mixed integer linear programming problem in each time slot, we analyze its inherent structure and propose an efficient algorithm to solve it based on Brenner's method. Our proposed algorithm makes online decisions rely only on the current system state and achieve cost emission trade‐off. Finally, the effectiveness of our algorithm is evaluated by extensive simulations based on real‐world data traces.

Virtual Network Functions Routing and Placement for Edge Cloud Latency Minimization

As a new way to design, deploy, and manage network services, network functions virtualization (NFV) decouples the network functions, from one or more physical network infrastructures and black boxes so they can run in software. It therefore comes as no surprise that NFV originated from service providers, who were looking to improve the deployment of new network services to support their revenue and growth objectives. Within the NFV ecosystem, high availability, and low latency are one of the key quality of service (QoS) benefits that service providers can expect from the 5G Cloud and the NFV networks to make delay-critical services such as remote surgery a reality. Therefore, network services should be placed, chained, and routed through the network considering users/tenants stringent QoS and service-level agreement requirements. To this end, routing and placement optimization plays a major role in improving network performance and the overall network cost. In this paper, we study the problem of virtual network functions (VNFs) placement and routing across the physical hosts to minimize overall latency defined as the queuing delay within the edge clouds and in network links. In that respect, this paper takes a holistic view by considering not only VNFs chaining and placement problem but also considering the flows routing aspect since these two problems are inter-related and have a major impact on network latency.

Resource management for deadline constrained MapReduce jobs for minimising energy consumption

Cloud computing has emerged as one of the leading platforms for processing large-scale data intensive applications. Such applications are executed in large clusters and data centres which require a substantial amount of energy. Energy consumption within data centres accounts for a considerable fraction of costs and is a significant contributor to global greenhouse gas emissions. Therefore, minimising energy consumption in data centres is a critical concern for data centre operators, cluster owners, and cloud service providers. In this paper, we devise a novel energy aware MapReduce resource manager for an open system, called EAMR-RM, that can effectively perform matchmaking and scheduling of MapReduce jobs each of which is characterised by a service level agreement (SLA) for performance that includes a client specified earliest start time, execution time, and a deadline with the objective of minimising data centre energy consumption. Performance analysis demonstrates that for a range of system and workload parameters experimented with the proposed technique can effectively satisfy SLA requirements while achieving up to a 45% reduction in energy consumption compared to approaches which do not consider energy in resource management decisions.

CRED: Cloud Right-Sizing with Execution Deadlines and Data Locality

As demands for cloud-based data processing continue to grow, cloud providers seek effective techniques that deliver value to the businesses without violating Service Level Agreements (SLAs). Cloud right-sizing has emerged as a very promising technique for making cloud services more cost-effective. In this paper, we present CRED, a novel framework for cloud right-sizing with execution deadlines and data locality constraints. CRED jointly optimizes data placement and task scheduling in data centers with the aim of minimizing the number of nodes needed while meeting users’ SLA requirements. We formulate CRED as an integer optimization problem and present a heuristic algorithm with provable performance guarantees to solve the problem. Competitive ratios of the proposed algorithm are quantified in closed form for arbitrary task parameters and cloud configurations. We also extend our work to obtain a resilient solution, which allows successful recovery at run time from any single node failure and is guaranteed to meet both deadline and locality constraints. Simulation results using Google trace show that our proposed algorithm significantly outperforms existing heuristics such as first-fit by reducing the number of required active servers by up to 47 percent, and achieves near-optimal performance. We also show that our algorithm can significantly improve utilization of both computational resources and storage space by up to 28 and 15 percent, respectively.

SERVICE LEVEL AGREEMENTS AND DEPENDABILITY

The Service Level Agreement (SLA) is an efficient and proven tool for regulation of the relations between the supplier and the user of services that is designed to ensure their quality. Such agreements are well known and successfully used in the information and communication industry. They are also applicable in other areas. Essentially, SLA stipulates certain requirements for the service level of which the fulfilment is guaranteed by the provider. In case of SLA violation the service provider is usually financially liable. As a rule, in such cases the user is remunerated with a discount for services provided in the following accounting period. Dependability requirements are an important part of the SLA. The purpose of this paper is to familiarize a wide range of experts from various industries with the general matters of SLA application and the aspects related to the dependability requirements specification. The paper refers to the relevant documents of international standardization organizations (ITU, ISO/IEC, ETSI, TMForum) and the Russian standards. Recommendations are given for selecting the dependability indicators and standard values to be included in the SLA, as well as for defining the amounts of compensation payed by service providers to the customers in case of non-compliance with requirements for the availability factor. The availability factor is normally used in the SLA as the primary dependability indicator that defines the allowable total time of non-operability over the specified base period. Additionally, a client might be interested in restricting the duration of each individual downtime as well. For that purpose, the guaranteed recovery time can also be specified and exceeding this time would be deemed an SLA violation. The choice of the standard values for inclusion in the SLA is a search for a compromise between the intent to satisfy the user requirements and the wish to get ahead of the competition on the one hand and the requirement to ensure the feasibility of the assumed obligations and minimize the risk of SLA violation that involve financial and reputational losses on the other hand. Therefore, before proposing an SLA to a customer, a service provider must thoroughly analyze its actual ability to make sure that the probability of SLA requirements violation is sufficiently low. The computational or computational and experimental methods are suggested for its evaluation. The amount of compensation for a violation depends on its gravity, i.e. the achieved and the standard values of an indicator. In practice, this relation is usually expressed with a step (piecewise constant) function. A formula is proposed that expresses the theoretical relation between the relative amount of compensation for violation of the availability factor requirements and the severity of violation and the standard value of this indicator. It can be used in defining the technically substantiated reference for SLA conditions development and assessment, of which the value will be relevant to both the service providers and users.

Energy-Spectral Efficiency Tradeoffs in 5G Multi-Operator Networks With Heterogeneous Constraints

Along with spectral efficiency (SE), energy efficiency (EE) is a key performance metric for the design of 5G and beyond 5G (B5G) wireless networks. At the same time, infrastructure sharing among multiple operators has also emerged as a new trend in wireless communication networks. This paper presents an optimization framework for EE and SE maximization in a network, where radio resources are shared among multiple operators. We define a heterogeneous service level agreement (SLA) framework for a shared network, in which the constraints of different operators are handled by two different multi-objective optimization approaches namely the utility profile and scalarization methods. Pareto-optimal solutions are obtained by merging these approaches with the theory of generalized fractional programming. The approach applies to both noise-limited and interference-limited systems, with single-carrier or multi-carrier transmission. Extensive numerical results illustrate the effect of the operator specific SLA requirements on the global spectral and EE. Three network scenarios are considered in the numerical results, each one corresponding to a different SLA, with different operator-specific EE and SE constraints.

ScHeduling of jobs and Adaptive Resource Provisioning (SHARP) approach in cloud computing

Affordability of appropriate computing resources for satisfying prerequisites of Service Level Agreement (SLA) of clients and optimal utilization of cloud service providers are limited in the present scenario of cloud computing. To overcome these limitations, researchers have exploited various scheduling algorithms to process the deadline based autonomous jobs. The scheduling algorithms however do not support multiprocessor demand and adaptive resource provisioning. This inference triggers to propose a new approach called ScHeduling of jobs and Adaptive Resource Provisioning (SHARP) in cloud computing to handle independent jobs that processes the jobs in a multilevel manner. The SHARP approach embeds multiple criteria decision analysis to preprocess the jobs, multiple attribute job scheduling to prioritize the jobs and adaptive resource provisioning to provide resources dynamically. These contributions alleviate SLA violations in terms of deadline, upgrade client satisfaction and enhance resource utilization. The empirical studies verify the proposed approach in a cloud environment and show the necessity of the proposed approach to support elastic resource provisioning and meet SLA requirements.

Deadline Guaranteed Service for Multi-Tenant Cloud Storage

It is imperative for cloud storage systems to be able to provide deadline guaranteed services according to service level agreements (SLAs) for online services. In spite of many previous works on deadline aware solutions, most of them focus on scheduling work flows or resource reservation in datacenter networks but neglect the server overload problem in cloud storage systems that prevents providing the deadline guaranteed services. In this paper, we introduce a new form of SLAs, which enables each tenant to specify a percentage of its requests it wishes to serve within a specified deadline. We first identify the multiple objectives (i.e., traffic and latency minimization, resource utilization maximization) in developing schemes to satisfy the SLAs. To satisfy the SLAs while achieving the multi-objectives, we propose a Parallel Deadline Guaranteed ( PDG ) scheme, which schedules data reallocation (through load re-assignment and data replication) using a tree-based bottom-up parallel process. The observation from our model also motivates our deadline strictness clustered data allocation algorithm that maps tenants with the similar SLA strictness into the same server to enhance SLA guarantees. We further enhance PDG in supplying SLA guaranteed services through two algorithms: i) a prioritized data reallocation algorithm that deals with request arrival rate variation, and ii) an adaptive request retransmission algorithm that deals with SLA requirement variation. Our trace-driven experiments on a simulator and Amazon EC2 show the effectiveness of our schemes for guaranteeing the SLAs while achieving the multi-objectives.

A Framework for Consumer-Centric SLA Management of Cloud-Hosted Databases

Service Level Agreements (SLA) represent the contract which captures the agreed upon guarantees between a service provider and its customers. The specifications of existing service level agreements (SLA) for cloud services are not designed to flexibly handle even relatively straightforward performance and technical requirements of consumer applications. In this article, we present a novel approach for SLA-based management of cloud-hosted databases from the consumer perspective. We present an end-to-end framework for consumer-centric SLA management of cloud-hosted databases. The framework facilitates adaptive and dynamic provisioning of the database tier of the software applications based on application-defined policies for satisfying their own SLA performance requirements, avoiding the cost of any SLA violation and controlling the monetary cost of the allocated computing resources. In this framework, the SLA of the consumer applications are declaratively defined in terms of goals which are subjected to a number of constraints that are specific to the application requirements. The framework continuously monitors the application-defined SLA and automatically triggers the execution of necessary corrective actions (scaling out/in the database tier) when required. The framework is database platform-agnostic, uses virtualization-based database replication mechanisms, and requires zero source code changes of the cloud-hosted software applications. The experimental results demonstrate the effectiveness of our SLA-based framework in providing the consumer applications with the required flexibility for achieving their SLA requirements.

A control‐theoretic approach towards joint admission control and resource allocation of cloud computing services

SummaryMeeting the performance specifications of consolidated web services in a data center is challenging, as the control of the underlying cloud computing infrastructure must meet the service level agreement requirements and satisfy the system's constraints. In this article, we address the admission control and resource allocation problem jointly, by establishing a unified modeling and control framework. Convergence to a desired reference point and stability and feasibility of the control strategy are guaranteed while achieving high performance of the co‐hosted web applications. The efficacy of the proposed approach is illustrated in a real test bed. Copyright © 2015 John Wiley & Sons, Ltd.

Shared execution of recurring workloads in MapReduce

With the increasing complexity of data-intensive MapReduce workloads, Hadoop must often accommodate hundreds or even thousands of recurring analytics queries that periodically execute over frequently updated datasets, e.g., latest stock transactions, new log files, or recent news feeds. For many applications, such recurring queries come with user-specified service-level agreements (SLAs), commonly expressed as the maximum allowed latency for producing results before their merits decay. The recurring nature of these emerging workloads combined with their SLA constraints make it challenging to share and optimize their execution. While some recent efforts on multi-job optimization in MapReduce have emerged, they focus on only sharing work among ad-hoc jobs on static datasets. Unfortunately, these sharing techniques neither take the recurring nature of the queries into account nor guarantee the satisfaction of the SLA requirements. In this work, we propose the first scalable multi-query sharing engine tailored for recurring workloads in the MapReduce infrastructure, called "Helix" . Helix deploys new sliced window-alignment techniques to create sharing opportunities among recurring queries without introducing additional I/O overheads or unnecessary data scans. And then, Helix introduces a cost/benefit model for creating a sharing plan among the recurring queries, and a scheduling strategy for executing them to maximize the SLA satisfaction. Our experimental results over real-world datasets confirm that Helix significantly outperforms the state-of-art techniques by an order of magnitude.

Energy-efficient resilient optical networks: Challenges and trade-offs

Energy efficiency and resilience are two well established research topics in optical transport networks. However, their overall objectives (i.e., power minimization and resource utilization/ availability maximization) conflict. In fact, provisioning schemes optimized for best resilience performance are in most cases not energy-efficient in their operations, and vice versa. However, very few works in the literature consider the interesting issues that may arise when energy efficiency and resilience are combined in the same networking solution. The objective of this article is to identify a number of research challenges and trade-offs for the design of energy-efficient and resilient optical transport networks from the perspective of long-term traffic forecasts, short-term traffic dynamics, and service level agreement requirements. We support the challenges with justifying numbers based on lessons learned from our previous work. The article also discusses suitable metrics for energy efficiency and resilience evaluation, in addition to a number of steps that need to be taken at the standardization level to incorporate energy efficiency into already existing and well established protocols.

Toward SLA-constrained service composition: An approach based on a fuzzy linguistic preference model and an evolutionary algorithm

In a market-oriented service computing environment, both back-end SLA (service level agreement) offers and front-end SLA requirements should be considered when performing service composition. In this paper, we address the optimization problem of SLA-constrained service composition and focus on the following issues: the difficulties related to preference definition and to weight assignment, the limitation of linear utility functions in identifying preferred skyline solutions, and the efficiency and scalability requirements of the optimization algorithm. We present a systematic approach based on a fuzzy preference model and on evolutionary algorithms. Specifically, we first model this multi-objective optimization problem using the weighted Tchebycheff distance rather than a linear utility function. We then present a fuzzy preference model for preference representation and weight assignment. In the model, a set of fuzzy linguistic preference terms and their properties are introduced for establishing consistent preference order of multiple QoS dimensions, and a weighting procedure is proposed to transform the preference into numeric weights. Finally, we present two evolutionary algorithms, i.e., single_EA and hybrid_EA, that implement different optimization objectives and that can be used in different SLA management scenarios for service composition. We conduct a set of experimental studies to evaluate the effectiveness of the proposed algorithms in determining the optimal solutions, and to evaluate their efficiency and scalability for different problem scales.

A design space for dynamic service level agreements in OpenStack

This paper does a systematic review of the possible design space for cloud-hosted applications that may have changing resource requirements that need to be supported through dynamic service level agreements (SLAs). The fundamental SLA functions are reviewed: Admission Control, Monitoring, SLA Evaluation, and SLA Enforcement - a classic autonomic control cycle. This is followed by an investigation into possible application requirements and SLA enforcement mechanisms. We then identify five basic Load Types that a dynamic SLA system must manage: Best Effort, Throttled, Load Migration, Preemption and Spare Capacity. The key to meeting application SLA requirements under changing surge conditions is to also manage the spare surge capacity. The use of this surge capacity could be managed by one of several identified load migration policies. A more detailed SLA architecture is presented that discusses specific SLA components. This is done in the context of the OpenStack since it is open source with a known architecture. Based on this SLA architecture, a research and development plan is presented wherein fundamental issues are identified that need to be resolved through research and experimentation. Based on successful outcomes, further developments are considered in the plan to produce a complete, end-to-end dynamic SLA capability. Executing on this plan will take significant resources and organization. The NSF Center for Cloud and Autonomic Computing is one possible avenue for pursuing these efforts. Given the growing importance of cloud performance management in the wider marketplace, the cloud community would be well-served to coordinate cloud SLA development across organizations such as the IEEE, Open Grid Forum, and the TeleManagement Forum.

A Cool Scheduler for Multi-Core Systems Exploiting Program Phases

Rapid growth of cloud computing services have led to creation of large scale enterprise data centers which consume great amounts of energy. Data centers usually have an service level agreement (SLA) between the clients and the service providers, which specify the terms and quality of service to be provided. In this paper, we consider a situation in a data center where multiple user applications are executing on a multi-core system and each application may have a specified SLA requirement. We design a voltage and frequency scheduler (the “cool” scheduler) that can be used in enterprise data centers to provide CPU energy saving under the specified SLA requirement by exploiting the applications’ run-time program phases. Our design greatly improves the computation efficiency compared to other recently published works. The scheduler is built into the Linux kernel and evaluated against SPEC CPU2006 and Phoronix Test Suite on a quad-core system. Experiment result demonstrates that our cool scheduler achieves 25.8% energy saving on average with 8.7% performance loss under the given SLA requirement (10% allowed performance loss). Our design achieves 35.8% and 31.6% more energy saving compared to two of the most advanced related works.

Efficient Resources Provisioning Based on Load Forecasting in Cloud

Cloud providers should ensure QoS while maximizing resources utilization. One optimal strategy is to timely allocate resources in a fine-grained mode according to application's actual resources demand. The necessary precondition of this strategy is obtaining future load information in advance. We propose a multi-step-ahead load forecasting method, KSwSVR, based on statistical learning theory which is suitable for the complex and dynamic characteristics of the cloud computing environment. It integrates an improved support vector regression algorithm and Kalman smoother. Public trace data taken from multitypes of resources were used to verify its prediction accuracy, stability, and adaptability, comparing with AR, BPNN, and standard SVR. Subsequently, based on the predicted results, a simple and efficient strategy is proposed for resource provisioning. CPU allocation experiment indicated it can effectively reduce resources consumption while meeting service level agreements requirements.