Geo-distributed Cloud Research Articles

Cost minimization approach with hybrid optimization based task scheduling in Geo-distributed cloud

ABSTRACT To manage the huge data, cloud computing necessitates a significant number of disc I/O, network bandwidth and CPU cycles. To handle the massive volumes of data, the programming paradigm known as data flow integrates the tasks to a graph structure. Task scheduling is one of the main areas of current study that primarily aims to allocate the tasks to resources. It is essential to organise the tasks in a way which saves time and makes the most use of available resources. In recent years, scholars have offered different job scheduling techniques. This work intends to introduce a novel cost minimisation for big data processing with respect to task allocation or scheduling. Initially, the big data is processed under Map Reduce framework. This allocation processes is taken as the optimisation issue, which is solved under the consideration of minimisation factors like execution cost, makespan, communication cost, and energy consumption and risk. For solving this issue, a new hybrid optimisation algorithm known as Hybrid bald eagle and Archimedes Optimisation algorithm (HBE-ArCA) is developed. The betterment of HBE-ArCA is proven regarding cost, makespan and so on.

Cost-efficient flow migration for SFC dynamical scheduling in geo-distributed clouds

Combining Network Function Virtualization (NFV) and Software Defined Network (SDN) enables the delivery of network services in a flexible and cost-efficient manner. NFV allows elasticity scheduling of Virtual Network Function (VNF) instances to adopt the dynamic variation of network flows, such as VNF instance scale out or in and load balancing. To implement the elasticity scheduling, existing flows need to be re-routed through the VNF instances. One major problem is how to select appropriate flows for migration to reduce migration cost and improve network performance after migration. In this paper, we propose a Cost-efficient Flow Migration framework, CFM, to jointly optimize flow migration cost and network performance after migration (including flow routing cost and maximum VNF instance load) for flow migration in Service Function Chain manner. Experimental results demonstrate that CFM can reduce flow routing cost after migration by 43%∼62% compared with previous works, while achieving similar flow migration cost in scale out/in scenario. In load balancing scenario, CFM can also reduce flow migration cost by 65%∼79% and flow routing cost after migration by 43%∼79% compared with previous works, while ensuring that the maximum VNF instance load does not significantly increase.

Achieving Cost Optimization for Tenant Task Placement in Geo-Distributed Clouds

Achieving Cost Optimization for Tenant Task Placement in Geo-Distributed Clouds

Energy and carbon-aware initial VM placement in geographically distributed cloud data centers

Energy and carbon efficiency has become one of the most important research topics in the cloud computing research area because of the growing demand for cloud data center infrastructures. In addition to the environmental impact of emitting carbon, local authorities impose a carbon tax based on the amount of emitted carbon by the data centers. Therefore, this paper introduces the energy and carbon-aware initial VM placement (ECAIVMP), one of the most efficient strategies for minimizing the total energy consumption and carbon emission of cloud data centers without sacrificing the quality of service. Our proposed algorithm considers multiple geo-distributed cloud data centers and acknowledges both IT and non-IT power usage during the virtual machine placement process. An extensive simulation in the CloudSim Plus framework is conducted to evaluate the performance of our proposed algorithm using real-world datasets, including workloads, power model benchmarks, and outside temperature. The experimental results demonstrate that our algorithm is able to reduce energy consumption by 17% and carbon emissions by 6% on average compared to other competitor algorithms.

Dual Optimization of Revenue and Expense in Geo-Distributed Data Centers Using Smart Grid

Exorbitant energy expenses can supersede data center profits. Electricity prices often vary across the geographic regions, caused by gaps in the supply-demand, time of use, and production cost factors. Geo-distributed cloud data centers facilitated by a smart grid and enabled by cloud computing can potentially utilize the spatiotemporal diversity of energy prices to reduce operational expenditure and maximize profit. In this paper, we solve the data center profit by formulating it as a constrained multi-objective optimization problem. The proposed solution utilizes an evolutionary algorithm-based higher-level heuristic that optimizes data center revenue and expense objectives simultaneously. The proposed technique provides system managers with trade-off solutions suited to varied operational scenarios. Ours is a multi-step approach, utilizing the optimization scheme to obtain Pareto optimal solutions for the request dispatch and resource allocation problem. When broadly evaluated against a comparative resource optimization scheme, our technique increases revenue while lowering expense and collectively yields a higher profit. It exhibits such performance over a broad range of price changes regardless of the data center's size and utilization level. The extensive simulation results ascertain the effectiveness of the proposed approach across a myriad of system parameters.

Adaptive and Scalable Caching With Erasure Codes in Distributed Cloud-Edge Storage Systems

Erasure codes have been widely used to enhance data resiliency with low storage overheads. However, in geo-distributed cloud storage systems, erasure codes may incur high service latency as they require end users to access remote storage nodes to retrieve data. An elegant solution to achieving low latency is to deploy caching services at the edge servers close to end users. In this paper, we propose adaptive and scalable caching schemes to achieve low latency in the cloud-edge storage system. Based on the measured data popularity and network latencies in real time, an adaptive content replacement scheme is proposed to update caching decisions upon the arrival of requests. Theoretical analysis shows that the reduced data access latency of the replacement scheme is at least 50% of the maximum reducible latency. With the low computation complexity of our design, nearly no extra overheads will be introduced when handling intensive data flows. For further performance improvements without sacrificing its efficiency, an adaptive content adjustment scheme is presented to replace the subset of cached contents that incur the aforementioned performance loss. Driven by real-world data traces, extensive experiments based on Amazon Simple Storage Service demonstrate the effectiveness and efficiency of our design.

Core-selecting auction-based mechanisms for service function chain provisioning and pricing in NFV markets

With the rapid development of network function virtualization (NFV), an ever-increasing number of enterprises, businesses and operators are resorting to network service provisioning by means of the service function chains (SFCs). Toward this trend, NFV markets are emerging, where users rent SFCs from the network service provider (NSP) to execute their tasks, and the NSP uses the computation resources across geo-distributed clouds to construct various SFCs. Such NFV markets need well-designed pricing mechanisms for SFCs since the users have motivations to strategize around their payments. The Vickrey–Clarke–Groves (VCG) mechanism is a natural fit here, which can ensure that truthful bidding is a dominant strategy. However, VCG mechanism is susceptible to the problems of collusion and shill bidding. To address these problems, in this paper, we design a core-selecting auction-based mechanism (CSAM) for SFC provisioning and pricing in the NFV market. This mechanism considers dynamic provisioning, flexible deployment and reservation prices across geo-distributed clouds. Specifically, we first formulate the core of NFV auction, and prove that an in-core outcome can prevent collusion and shill bidding. Next, we use a couple of correlated linear program (LP) and quadratic program (QP) to design an in-core VCG-nearest payment algorithm, aiming to acquire an in-core outcome and minimize users’ incentives to bid untruthfully. Simulation results verify the strict theoretical analysis, and validate the effectiveness and efficiency of our proposed CSAM.

A Survey on Machine Learning for Geo-Distributed Cloud Data Center Management

Cloud workloads today are typically managed in a distributed environment and processed across geographically distributed data centers. Cloud service providers have been distributing data centers globally to reduce operating costs while also improving quality of service by using intelligent workload and resource management strategies. Such large scale and complex orchestration of software workload and hardware resources remains a difficult problem to solve efficiently. Researchers and practitioners have been trying to address this problem by proposing a variety of cloud management techniques. Mathematical optimization techniques have historically been used to address cloud management issues. But these techniques are difficult to scale to geo-distributed problem sizes and have limited applicability in dynamic heterogeneous system environments, forcing cloud service providers to explore intelligent data-driven and Machine Learning (ML) based alternatives. The characterization, prediction, control, and optimization of complex, heterogeneous, and ever-changing distributed cloud resources and workloads employing ML methodologies have received much attention in recent years. In this article, we review the state-of-the-art ML techniques for the cloud data center management problem. We examine the challenges and the issues in current research focused on ML for cloud management and explore strategies for addressing these issues. We also discuss advantages and disadvantages of ML techniques presented in the recent literature and make recommendations for future research directions.

An Intelligent Resource Reservation for Crowdsourced Live Video Streaming Applications in Geo-Distributed Cloud Environment

Crowdsourced live video streaming (livecast) services such as Facebook Live, YouNow, Douyu, and Twitch are gaining more momentum recently. Allocating the limited resources in a cost-effective manner while maximizing the quality of service (QoS) through real-time delivery and the provision of the appropriate representations for all viewers is a challenging problem. In our article, we introduce a machine-learning based predictive resource allocation framework for geo-distributed cloud sites, considering the delay and quality constraints to guarantee the maximum QoS for viewers and the minimum cost for content providers. First, we present an offline optimization that decides the required transcoding resources in distributed regions near the viewers with a tradeoff between the QoS and the overall cost. Second, we use machine learning to build forecasting models that proactively predict the approximate transcoding resources to be reserved at each cloud site ahead of time. Finally, we develop a greedy nearest and cheapest algorithm (GNCA) to perform the resource allocation of real-time broadcasted videos on the rented resources. Extensive simulations have shown that GNCA outperforms the state-of-the-art resource allocation approaches for crowdsourced live streaming by achieving more than 20% gain in terms of system cost while serving the viewers with relatively lower latency.

Fault-tolerant scheduling and data placement for scientific workflow processing in geo-distributed clouds

Scientific workflow processing in geo-distributed cloud is crucial for the scheduling of large-scale tasks and the massive data placement among tasks. However, the task execution time and energy consumption of data transmission are two urgent issues when a scientific workflow is processed in the different geo-distributed data centers. Aiming at the data placement problem, this paper proposes a Lagrangian relaxation method. This method considers the workload balance, storage capacity, data dependency, transmission bandwidth, and transmission cost to obtain the minimum time of data transmission time. Aiming at the task scheduling problems, a fault-tolerant scheduling strategy is proposed. The strategy optimizes the task scheduling mechanism by considering the task execution time and energy consumption. Finally, the performance of the proposed methods is evaluated via extensive experiments In terms of the data placement, the experiment results imply that the data transmission time of the proposed relaxation algorithm can averagely achieve up to 14.61%, 38.03%, and 39.57% reduction of over that of ILP-FDP algorithm, GA-DPSO algorithm, and GPDP algorithm, respectively. As for the fault-tolerant scheduling, the energy consumption of the TSPT algorithm is the lowest. Compared with the MTS algorithm, EODS algorithm and EWTS algorithm, the average gains of the proposed algorithm are 15.33%, 16.65%, and 28.96%, respectively. Compared with the benchmark algorithms, the task execution time of the TSPT algorithm can averagely reduce up to 12.78,18.85 and 25.65, respectively.

Multi-objective optimization of data deployment and scheduling based on the minimum cost in geo-distributed cloud

Nowadays, the amount of data coming from terminal devices is increasing on a vast scale. Data centers in different geographical locations store vast amounts of data. In this paper, to optimize the data deployment problem, a data layout algorithm that can be satisfied with the load balancing of the cluster is introduced. Under the three constraints of cost, capacity, and load balance, this paper presents the idea of ant colony optimization and uses the Lagrangian relaxation method to verify the value of the optimal solution. For improving the execution efficiency of various cloud data centers, a task scheduling method is proposed. This method solves the problems of delay and transmission cost and then uses the gray Markov-based prediction method to predict the dynamic changes of resources of different cluster nodes and select the most suitable task scheduling node. As shown in the experiment, the completion time of the application can be shortened by this algorithm. Meanwhile, in this paper, the consumption of cluster resources is reduced, and the throughput of the cluster is improved.

Efficient multi-attribute precedence-based task scheduling for edge computing in geo-distributed cloud environment

Efficient multi-attribute precedence-based task scheduling for edge computing in geo-distributed cloud environment

Optimal Data Placement and Replication Approach for SIoT with Edge

Social networks (SNs) are sources with extreme number of users around the world who are all sharing data like images, audio, and video to their friends using IoT devices. This concept is the so-called Social Internet of Things (SIot). The evolving nature of edge-cloud computing has enabled storage of a large volume of data from various sources, and this task demands an efficient storage procedure. For this kind of large volume of data storage, the usage of data replication using edge with geo-distributed cloud service area is suited to fulfill the user’s expectations with low latency. The major issue is the way to store the data and replicate these large data items optimally and allocate the request from the data center efficiently. For efficient storage of these data, we use edge server, which is part of the cloud server, in this study. Thus, the data are distributed and stored with quick access, which will reduce the latency with response. The proposed data placement approach learns with machine learning (ML) algorithm called radial basis kernel function assisted with support vector machine (RBF-SVM) to classify the data center for storing the user and friend’s data from the SIoT devices. These learning algorithms will be used to predict the workload of the data stored in the data center as either edge or cloud depending on the existing time slots. The data placement with dynamic nature is also optimized using the proposed dynamic graph partitioning (GP) method to meet the individual user’s demand of low latency with minimum costs. This way will keep the SIoT data placement efficient and effective over time. Accordingly, this proposed data placement and replication approach introduces three kinds of innovations compared with the existing data placement approach. (i) Rather than storing the user data in a single cloud, this study uses the edge server closest to the SIoT devices for faster access with reduced response time. (ii) The classification algorithm called RBF-SVM is used to find storage for user for reducing data replication. (iii) Dynamic GP is introduced for data placement with reduced latency and minimum cost to fulfil the dynamic nature of the SN. The simulation result of this approach obtains reduced latency of 130 ms and minimum cost compared with those of the existing data placement approaches. Therefore, our proposed data placement with ML-based learning on edge provides promising results in terms of efficiency, effectiveness, and performance with reduced latency and minimum cost.

LECC: Location, energy, carbon and cost-aware VM placement model in geo-distributed DCs

The wide adoption of the cloud computing model in the business environment has led to a rapid increase in the development of geo-distributed data centers (DCs) to support customers’ needs. On the other hand, cloud DCs that contain thousands of computing and storage nodes consumes a large amount of energy that leads to a high carbon footprint. Therefore, minimizing the geo-distributed DCs’ energy consumption is a must which will decrease the cloud providers’ operational cost and minimize the high non-environment carbon emission. In addition, minimizing the cloud users’ network latency, in such a distributed environment, is one of the important challenges faced by cloud providers. Thus, to address these challenges, this paper proposed a novel location, energy, carbon, and cost-aware (LECC) virtual machine (VM) placement model for geo-distributed cloud DCs. Both online and offline placement problems are tackled. The migration technology is employed in consolidating the VMs to less number of active servers for significant energy reductions. The VM placement problem is formulated as a multi-objective optimization problem and solved by greedy policies. Also, an intelligent machine-learning model is constructed and implemented to leverage the performance of the LECC model. To validate the usefulness of the LECC model, extensive simulations using synthetic and real data are conducted on CloudSim toolkit. The experimental results showed the merits of our proposed LECC model in solving the important VM placement problem.

Adaptive priority-based data placement and multi-task scheduling in geo-distributed cloud systems

With the rapid development and the widespread use of cloud computing in various applications, the number of users distributed in different regions has grown exponentially. Therefore, the Geo-distributed cloud systems have become a research hotspot and big data processing technology has also emerged. Nowadays, the most widely used big data processing framework is Spark. However, massive amounts of data are generated every moment, and the processing procedure becomes more and more complex, the execution efficiency of Spark has been greatly affected. In the Spark frame of geo-distributed cloud systems, aiming at the data placement problem, the data placement strategy based on RDD dynamic weight is introduced. The target node is selected with a strong computation capacity to place the data. Aiming at the problems of multi-task scheduling, a task will be scheduled to a node whose computation capacity can satisfy the requirement of this task. And then considering job classification and computing node performance, the optimized task scheduling strategy is in traduced. Experiments show that our algorithms can effectively adjust the weight of node data placement according to the actual task execution information, and shorten the task execution time.

Optimized Contract-Based Model for Resource Allocation in Federated Geo-Distributed Clouds

In the era of Big Data, with data growing massively in scale and velocity, cloud computing and its pay-as-you-go model continues to provide significant cost benefits and a seamless service delivery model for cloud consumers. The evolution of small-scale and large-scale geo-distributed datacenters operated and managed by individual Cloud Service Providers (CSPs) raises new challenges in terms of effective global resource sharing and management of autonomously-controlled individual datacenter resources towards a globally efficient resource allocation model. Earlier solutions for geo-distributed clouds have focused primarily on achieving global efficiency in resource sharing, that although tries to maximize the global resource allocation, results in significant inefficiencies in local resource allocation for individual datacenters and individual cloud provi ders leading to unfairness in their revenue and profit earned. In this paper, we propose a new contracts-based resource sharing model for federated geo-distributed clouds that allows CSPs to establish resource sharing contracts with individual datacenters apriori for defined time intervals during a 24 hour time period. Based on the established contracts, individual CSPs employ a contracts cost and duration aware job scheduling and provisioning algorithm that enables jobs to complete and meet their response time requirements while achieving both global resource allocation efficiency and local fairness in the profit earned. The proposed techniques are evaluated through extensive experiments using realistic workloads generated using the SHARCNET cluster trace. The experiments demonstrate the effectiveness, scalability and resource sharing fairness of the proposed model.

Federated Geo-Distributed Clouds: Optimizing Resource Allocation Based on Request Type Using Autonomous and Multi-objective Resource Sharing Model

Abstract Due to the problems exist in non-geographic federated clouds, the geographic ones are considered. Nevertheless, the approaches that have already been proposed to allocate resources across the geographical federated clouds have two basic problems that we will address in this article: (1) Lack of proper distribution of user requests leading to increases file transfer volume and cost, as well as response time to user requests, (2) Lack of appropriate resource sharing among requests due to: (1) the use of a centralized DC and (2) considering the satisfaction of single objective which case (1) suffers the problem of single-point of failure and case (2) raises an obstacle for the situations need considering multi conflicting objectives. Concerning the problem of one, it should be said that as federal DCs are distributed globally in the geographic clouds, the cost of file transfer between DCs in these clouds is more focused than the concentrated ones. Since there has been no work in this field in the geo-distributed federated clouds, we have presented a new scheduling mechanism based on hypervolume for the distribution of applications that leads to increasing service quality and reducing file transfer cost. Concerning the problem of two, the previous solutions in the geographic federated clouds have focused on a centralized resource sharing with single objective (increase of the cloud service provider (CSP) profit). These solutions not only just consider the CSP profitability, but, because of the possibility of failure of central broker of resource-sharing, suffer the single-point of failure. In this paper, we propose a new, autonomic and peer-to-peer multi-objective resource sharing approach that considers objectives: (1) enhancing the CSP's profit, (2) decreasing the network latency and (3) decreasing file transfer traffic and (3) increasing fairness in CSPs' profit. The techniques presented in this paper are evaluated by extensive experiments using real workloads. To validate the proposed method, we have extended the CloudSim tool. The results of our experiments show the increase of performance in the scheduling and resource-sharing objectives among which the main objectives of average rate of success, profit and execution time were enhanced 8.5%, 15.47% and 25.84%, respectively compared with previous studies.

Less is More: Service Profit Maximization in Geo-Distributed Clouds

Nowadays cloud providers purchase a good deal of bandwidth from Internet service providers to satisfy the growing requests from corporate customers for the exclusive use of inter-datacenter bandwidth. For exclusive bandwidth services, neither maximizing the revenue nor minimizing the cost can bring the maximal profit to cloud providers. The diversity of bandwidth prices and the random arrival time of user requests further increase the difficulty in economically scheduling the services to meet user requests from cloud providers. In this article, we propose to help cloud providers maximize their service profits by properly selecting user requests to serve rather than satisfying them all. We formulate the problem of service profit maximization and prove its NP-hardness. To handle offline request submission, we propose a solution that maximizes the service profit by alternately maximizing the service revenue and minimizing the service cost. To maximize service profit under online request submission, we propose an online scheduling algorithm that carefully handles the risk of not being able to pay off the incremental service cost and makes scheduling decisions in real time. Our extensive evaluations demonstrate that our solutions can achieve more than 1.6x the service profits of existing solutions.

Scalable and Adaptive Data Replica Placement for Geo-Distributed Cloud Storages

In geo-distributed cloud storage systems, data replication has been widely used to serve the ever more users around the world for high data reliability and availability. How to optimize the data replica placement has become one of the fundamental problems to reduce the inter-node traffic and the system overhead of accessing associated data items. In the big data era, traditional solutions may face the challenges of long running time and large overheads to handle the increasing scale of data items with time-varying user requests. Therefore, novel offline community discovery and online community adjustment schemes are proposed to solve the replica placement problem in a scalable and adaptive way. The offline scheme can find a replica placement solution based on the average read/write rates for a certain period of time. The scalability can be achieved as 1) the computation complexity is linear to the amount of data items and 2) the data-node communities can evolve in parallel for a distributed replica placement. Furthermore, the online scheme is adaptive to handle the bursty data requests, without the need to completely override the existing replica placement. Driven by real-world data traces, extensive performance evaluations demonstrate the effectiveness of our design to handle large-scale datasets.

RL-OPRA: Reinforcement Learning for Online and Proactive Resource Allocation of crowdsourced live videos

With the advancement of rich media generating devices, the proliferation of live Content Providers (CP), and the availability of convenient internet access, crowdsourced live streaming services have witnessed unexpected growth. To ensure a better Quality of Experience (QoE), higher availability, and lower costs, large live streaming CPs are migrating their services to geo-distributed cloud infrastructure. However, because of the dynamics of live broadcasting and the wide geo-distribution of viewers and broadcasters, it is still challenging to satisfy all requests with reasonable resources. To overcome this challenge, we introduce in this paper a prediction driven approach that estimates the potential number of viewers near different cloud sites at the instant of broadcasting. This online and instant prediction of distributed popularity distinguishes our work from previous efforts that provision constant resources or alter their allocation as the popularity of the content changes. Based on the derived predictions, we formulate an Integer-Linear Program (ILP) to proactively and dynamically choose the right data center to allocate exact resources and serve potential viewers, while minimizing the perceived delays. As the optimization is not adequate for online serving, we propose a real-time approach based on Reinforcement Learning (RL), namely RL-OPRA, which adaptively learns to optimize the allocation and serving decisions by interacting with the network environment. Extensive simulation and comparison with the ILP have shown that our RL-based approach is able to present optimal results compared to heuristic-based approaches.