Cluster Of Virtual Machines Research Articles

Efficient Task Scheduling in Cloud Computing using Multi-objective Hybrid Ant Colony Optimization Algorithm for Energy Efficiency

The efficiency of Internet services is determined by the Cloud computing process. Various challenges in computing are being faced, such as security, the efficient allocation of resources, which in turn results in the waste of resources. Researchers have explored a number of approaches over the past decade to overcome these challenges. The main objective of this research is to explore the task scheduling of cloud computing using multi-objective hybrid Ant Colony Optimization (ACO) with Bacterial Foraging (ACOBF) behavior. ACOBF technique maximized resource utilization (Service Provider Profit) and also reduced Makespan and user wait times Job request. ACOBF classifies the user job request in three classes based on the sensitivity of the protocol associated with each request, Schedule Job request in each class based on job request deadline and create a Virtual Machine (VM) cluster to minimize energy consumption. Based on comprehensive experimentation, the simulated results show that the performance of ACOBF outperforms the benchmarked techniques in terms of convergence, diversity of solutions and stability.

Experimental Setup for Investigating the Efficient Load Balancing Algorithms on Virtual Cloud

Cloud computing has emerged as the primary choice for developers in developing applications that require high-performance computing. Virtualization technology has helped in the distribution of resources to multiple users. Increased use of cloud infrastructure has led to the challenge of developing a load balancing mechanism to provide optimized use of resources and better performance. Round robin and least connections load balancing algorithms have been developed to allocate user requests across a cluster of servers in the cloud in a time-bound manner. In this paper, we have applied the round robin and least connections approach of load balancing to HAProxy, virtual machine clusters and web servers. The experimental results are visualized and summarized using Apache Jmeter and a further comparative study of round robin and least connections is also depicted. Experimental setup and results show that the round robin algorithm performs better as compared to the least connections algorithm in all measuring parameters of load balancer in this paper.

A Scalable Virtualized Server Cluster Providing Sensor Data Storage and Web Services

With the rapid development of the Internet of Things (IoT) technology, diversified applications deploy extensive sensors to monitor objects, such as PM2.5 air quality monitoring. The sensors transmit data to the server periodically and continuously. However, a single server cannot provide efficient services for the ever-growing IoT devices and the data they generate. This study bases on the concept of symmetry of architecture and quantities in system design and explores the load balancing issue to improve performance. This study uses the Linux Virtual Server (LVS) and virtualization technology to deploy a virtual machine (VM) cluster. It consists of a front-end server, also a load balancer, to dispatch requests, and several back-end servers to provide services. These receive data from sensors and provide Web services for browsing real-time sensor data. The Hadoop Distributed File System (HDFS) and HBase are used to store the massive amount of received sensor data. Because load-balancing is critical for resource utilization, this study also proposes a new load distribution algorithm for VM-based server clusters that simultaneously provide multiple services, such as sensor services and Web service. It considers the aggregate load of all back-end servers on the same physical server that provides multiple services. It also considers the difference between physical machines and VMs. Algorithms such as those for LVS, which do not consider these factors, can cause load imbalance between physical servers. The experimental results demonstrate that the proposed system is fault tolerant, highly scalable, and offers high availability and high performance.

Performance consequence of user space file systems due to extensive CPU sharing in virtual environment

Recently, FUSE based user space file systems have gained importance due to their ease of implementation. Their lower performance versus implementation benefits, compared to traditional in-kernel file systems, has always been a point of debate among researchers. As FUSE requires additional context switching to perform file related operations, there is a noticeable increase in CPU utilization. In this era of cloud computing, where the focus is shifting on running applications in virtual machines, increased CPU utilization during file operations can have considerable impact on performance, as resources of the hypervisor are shared among virtual machines. This degradation in performance can become a major cause of concern especially when resources are over-committed. There has been ongoing research regarding improvement in performance of user space file systems, and this will help in providing a systematic study on evaluating their performance. For an in-depth examination, we analyzed the performance of FUSE based file systems running in a guest virtual machine and highlighted some of the scenarios, where there could be a major damage to the file system’s performance. We have also showed here, that a careful selection of parameters like block sizes and type of read/write operations can improve performance of these file systems significantly by 40% to 45% even in heavily loaded environment. This study will inspire users and developers to enhance the performance of FUSE based file systems which are now commonly used in cluster of virtual machines.

Fine-grained data-locality aware MapReduce job scheduler in a virtualized environment

Big data overwhelmed industries and research sectors. Reliable decision making is always a challenging task, which requires cost-effective big data processing tools. Hadoop MapReduce is being used to store and process huge volume of data in a distributed environment. However, due to huge capital investment and lack of expertise to set up an on-premise Hadoop cluster, big data users seek cloud-based MapReduce service over the Internet. Mostly, MapReduce on a cluster of virtual machines is offered as a service for a pay-per-use basis. Virtual machines in MapReduce virtual cluster reside in different physical machines and co-locate with other non-MapReduce VMs. This causes to share IO resources such as disk and network bandwidth, leading to congestion as most of the MapReduce jobs are disk and network intensive. Especially, the shuffle phase in MapReduce execution sequence consumes huge network bandwidth in a multi-tenant environment. This results in increased job latency and bandwidth consumption cost. Therefore, it is essential to minimize the amount of intermediate data in the shuffle phase rather than supplying more network bandwidth that results in increased service cost. Considering this objective, we extended multi-level per node combiner for a batch of MapReduce jobs to improve makespan. We observed that makespan is improved up to 32.4 $$\%$$ by minimizing the number of intermediate data in shuffle phase when compared to classical schedulers with default combiners.

Cosmos : A Unified Accounting System both for the HTCondor and Slurm Clusters at IHEP

HTCondor was adopted to manage the High Throughput Computing (HTC) cluster at IHEP in 2016. In 2017 a Slurm cluster was set up to run High Performance Computing (HPC) jobs. To provide accounting services for these two clusters, we implemented a unified accounting system named Cosmos. Multiple workloads bring different accounting requirements. Briefly speaking, there are four types of jobs to account. First of all, 30 million single-core jobs run in the HTCondor cluster every year. Secondly, Virtual Machine (VM) jobs run in the legacy HTCondor VM cluster. Thirdly, parallel jobs run in the Slurm cluster, and some of these jobs are run on the GPU worker nodes to accelerate computing. Lastly, some selected HTC jobs are migrated from the HTCondor cluster to the Slurm cluster for research purposes. To satisfy all the mentioned requirements, Cosmos is implemented with four layers: acquisition, integration, statistics and presentation. Details about the issues and solutions of each layer will be presented in the paper. Cosmos has run in production for two years, and the status shows that it is a well-functioning system, also meets the requirements of the HTCondor and Slurm clusters.

CLOSURE: A cloud scientific workflow scheduling algorithm based on attack–defense game model

The multi-tenant coexistence service mode makes the cloud-based scientific workflow encounter the risks of being intruded. For this problem, we propose a CLoud scientific wOrkflow SchedUling algoRithm based on attack–defensE game model (CLOSURE). In the algorithm, attacks based on different operating system vulnerabilities are regarded as different “attack” strategies; and different operating system distributions in a virtual machine cluster executing the workflows are regarded as different “defense” strategies. The information of the attacker and defender is not balanced. In other words, the defender cannot obtain the information about the attacker’s strategies, while the attacker can acquire information about the defender’s strategies through a network scan. Therefore, we propose to dynamically switch the defense strategies during the workflow execution, which can weaken the network scan effects and transform the workflow security problem into an attack–defense game problem. Then, the probability distribution of the optimal mixed defense strategies can be achieved by calculating the Nash Equilibrium in the attack–defense game model. Based on this probability, diverse VMs are provisioned for workflow execution. Furthermore, a task-VM mapping algorithm based on dynamic Heterogeneous Earliest Finish Time (HEFT) is presented to accelerate the defense strategy switching and improve workflow efficiency. The experiments are conducted on both simulation and actual environment, experimental results demonstrate that compared with other algorithms, the proposed algorithm can reduce the attacker’s benefits by around 15.23%, and decrease the time costs of the algorithm by around 7.86%.

Dynamic ranking-based MapReduce job scheduler to exploit heterogeneous performance in a virtualized environment

“More data, more information.” Big data helps businesses and research communities to gain insights and increase productivity. Many public cloud service providers offer Hadoop MapReduce as a service based on pay-per-use via infrastructure as a service on clusters of virtual machines promising on-demand horizontal scaling. These clusters of virtual machines are launched in various physical machines across racks in cloud data centers. Such multi-tenancy negatively introduces performance heterogeneity for Hadoop virtual machines due to hardware heterogeneity and interference from co-located virtual machine. Performance heterogeneity largely affects MapReduce job latency and resource utilization of rented Hadoop virtual clusters. Default MapReduce schedulers assign map/reduce tasks assuming the hardware is homogeneous. Interference-aware schedulers perform by only observing the interference pattern generated by co-located virtual machines. These schedulers do not consider the heterogeneous performance of virtual machines. Therefore, we propose a dynamic ranking-based MapReduce job scheduler that places the map and reduces tasks based on a virtual machine’s performance rank to minimize job latency and improve resource utilization. Our proposed approach calculates the performance score for each virtual machine based on hardware heterogeneity and co-located virtual machine interference. Then, it ranks the virtual machines based on the map and reduce performance separately to place map and reduce tasks. To demonstrate our ideas, we have set a test bed with 29 virtual machines on eight physical machines with different configurations and capacities. We modify a default fair scheduler in Hadoop 2.x to incorporate our ideas and evaluate them with different workloads on the PUMA dataset. The proposed method is then compared against a default fair scheduler (resource-aware) and an interference-aware scheduler based on job latency and resource utilization. Finally, we argue in favor of our approach as it improves resource utilization by 30–65% and overall job latency by up to 30%.

Load prediction analysis based on virtual machine execution time using optimal sequencing algorithm in cloud federated environment

Virtual machine (VM) prediction and an effective resource management are the attractive areas in the cloud environment. VM prediction is an important task to execute the jobs for delay minimization and unnecessary states avoidance. Cloud computing attracted towards the increase in a number of applications that run on remote servers in parallel manner. Increase in parallelism reduces the CPU utilization adversely. Hence, the proper VM prediction and management are necessary stages in provisioning scheme. Also time required for allocating jobs is more in existing algorithms due to the number of computations involved. Therefore a novel algorithm is required to improve the performance of the job allocation with makespan reduction. In this paper the new algorithm is proposed that includes the VM capacity and execution time for load prediction and performance improvement purpose. Our proposed research work utilizes the VM clustering and optimization algorithms to improve job sequencing performance. The cost computation prior to clustering includes the VM capacity as a major factor. Clustering of VM with high-cost and isolation of low-cost and high-cost clusters reduces the searching time of VM and solve the imbalance state problem in traditional methods. The optimization algorithm with suitable initialization function reduces the time and steps for selection of VM for suitable job. The proposed model outperformance is established by the selected parameters.

Monte Carlo verification of radiotherapy treatments with CloudMC

BackgroundA new implementation has been made on CloudMC, a cloud-based platform presented in a previous work, in order to provide services for radiotherapy treatment verification by means of Monte Carlo in a fast, easy and economical way. A description of the architecture of the application and the new developments implemented is presented together with the results of the tests carried out to validate its performance.MethodsCloudMC has been developed over Microsoft Azure cloud. It is based on a map/reduce implementation for Monte Carlo calculations distribution over a dynamic cluster of virtual machines in order to reduce calculation time. CloudMC has been updated with new methods to read and process the information related to radiotherapy treatment verification: CT image set, treatment plan, structures and dose distribution files in DICOM format. Some tests have been designed in order to determine, for the different tasks, the most suitable type of virtual machines from those available in Azure. Finally, the performance of Monte Carlo verification in CloudMC is studied through three real cases that involve different treatment techniques, linac models and Monte Carlo codes.ResultsConsidering computational and economic factors, D1_v2 and G1 virtual machines were selected as the default type for the Worker Roles and the Reducer Role respectively. Calculation times up to 33 min and costs of 16 € were achieved for the verification cases presented when a statistical uncertainty below 2% (2σ) was required. The costs were reduced to 3–6 € when uncertainty requirements are relaxed to 4%.ConclusionsAdvantages like high computational power, scalability, easy access and pay-per-usage model, make Monte Carlo cloud-based solutions, like the one presented in this work, an important step forward to solve the long-lived problem of truly introducing the Monte Carlo algorithms in the daily routine of the radiotherapy planning process.

The trusted virtual machine cluster construction method based on particle wave equation

In current cloud computing platforms, the study of the security of the virtual machine in cloud computing has not considered the cloud environment, which is a complex clustering, and it also has not considered the fact that the virtual machine cluster resides on a single virtual machine. In order to solve the above problems, a method to build a trusted virtual machine groups based on the particle wave equation is proposed to build a trusted virtual machine group. Compared with the traditional security policy for a single virtual machine, the measurement parameters of our method can be formulated according to different environments, it has a better environmental adaptability, and by constructing a virtual machine cluster it can effectively reduce the security threats to virtual machines.

The trusted virtual machine cluster construction method based on particle wave equation

In current cloud computing platforms, the study of the security of the virtual machine in cloud computing has not considered the cloud environment, which is a complex clustering, and it also has not considered the fact that the virtual machine cluster resides on a single virtual machine. In order to solve the above problems, a method to build a trusted virtual machine groups based on the particle wave equation is proposed to build a trusted virtual machine group. Compared with the traditional security policy for a single virtual machine, the measurement parameters of our method can be formulated according to different environments, it has a better environmental adaptability, and by constructing a virtual machine cluster it can effectively reduce the security threats to virtual machines.

A survey and taxonomy of resource optimisation for executing bag-of-task applications on public clouds

Cloud computing has been widely adopted due to the flexibility in resource provisioning and on-demand pricing models. Entire clusters of Virtual Machines (VMs) can be dynamically provisioned to meet the computational demands of users. However, from a user’s perspective, it is still challenging to utilise cloud resources efficiently. This is because an overwhelmingly wide variety of resource types with different prices and significant performance variations are available.This paper presents a survey and taxonomy of existing research in optimising the execution of Bag-of-Task applications on cloud resources. A BoT application consists of multiple independent tasks, each of which can be executed by a VM in any order; these applications are widely used by both the scientific communities and commercial organisations. The objectives of this survey are as follows: (i) to provide the reader with a concise understanding of existing research on optimising the execution of BoT applications on the cloud, (ii) to define a taxonomy that categorises current frameworks to compare and contrast them, and (iii) to present current trends and future research directions in the area.

Piccolo: A Fast and Efficient Rollback System for Virtual Machine Clusters

Rollback is an effective technique to resume the system execution from a recorded intermediate state upon failures, without having to restart the entire system. However, in virtualized environments, rollback of a virtual machine cluster (VMC) produces high network traffic and long service disruption, particularly for a large cluster used for scientific computing, thereby imposing significant overhead both on network and applications. This paper proposes Piccolo, a fast and efficient rollback system, to restore a VMC from snapshot files over data center network. First, we exploit the similarity among VMC snapshots and leverage multicast to deliver the identical pages across VMs placed on disperse hosts, thereby bypassing unnecessary transmission of a large number of pages. Second, we analyze the impact on network traffic of varying VM placements in data center network, formulate the traffic aware placement as an optimization problem, and design a two-tier approximation algorithm that efficiently solves the problem. In addition to presenting Piccolo, we detail its implementation, and evaluate it by a set of experiments. The results show that Piccolo could achieve a significant reduction in terms of total sent data, network traffic and rollback latency compared to the existing generic techniques.

<sc>ATOM</sc>: Efficient Tracking, Monitoring, and Orchestration of Cloud Resources

The emergence of Infrastructure as a Service framework brings new opportunities, which also accompanies with new challenges in auto scaling, resource allocation, and security. A fundamental challenge underpinning these problems is the continuous tracking and monitoring of resource usage in the system. In this paper, we present ATOM , an efficient and effective framework to automatically track, monitor, and orchestrate resource usage in an Infrastructure as a Service (IaaS) system that is widely used in cloud infrastructure. We use novel tracking method to continuously track important system usage metrics with low overhead, and develop a Principal Component Analysis (PCA) based approach to continuously monitor and automatically find anomalies based on the approximated tracking results. We show how to dynamically set the tracking threshold based on the detection results, and further, how to adjust tracking algorithm to ensure its optimality under dynamic workloads. Lastly, when potential anomalies are identified, we use introspection tools to perform memory forensics on VMs guided by analyzed results from tracking and monitoring to identify malicious behavior inside a VM. We demonstrate the extensibility of ATOM through virtual machine (VM) clustering. The performance of our framework is evaluated in an open source IaaS system.

Improving performance by network-aware virtual machine clustering and consolidation

Modern data center consists of thousands of servers, racks and switches. Complicated structure means it requires well-designed algorithms to utilize resources of data centers efficiently. Current virtual machine scheduling algorithms mainly focus on the initial allocation of virtual machines based on the CPU, memory and network bandwidth requirements. However, when tasks finished or lease expired, related virtual machines would be deleted from the system which would generate resource fragments. Such fragments lead to unbalanced resource utilization and decline of communication performance. This paper investigates the network influence on typical applications in data centers and proposed a self-adaptive network-aware virtual machine clustering and consolidation algorithm to maintain an optimal system-wide status. Our consolidation algorithm periodically checks whether consolidation is necessary and then clusters and consolidates virtual machines to lower communication cost with an online heuristic. We used two benchmarks in a real environment to examine network influence on different tasks. To evaluate the advantages of the proposed algorithm, we also built a cloud computing testbed. Real workload trace-driven simulations and testbed-based experiments showed that, our algorithm greatly shortened the average finish time of map-reduce tasks and reduced time delay of web applications. Simulation results showed that our algorithm considerably reduced the amount of high-delay jobs, lowered the average traffic passed through aggregate switches and improved the communication ability among virtual machines.

Configuration Guidance Framework for Molecular Dynamics Simulations in Virtualized Clusters

With the advancement of cloud computing, there has been a growing interest in exploiting demand-based cloud resources for parallel scientific applications. To satisfy different needs for computing resources, cloud providers provide many different types of virtual machines (VMs) with various numbers of computing cores and amounts of memory. The cost and execution time of a scientific application vary depending on the types of VMs, number of VMs, and current status of the cloud due to interference among VMs. However, currently, cloud users are solely responsible for selecting the most effective VM configuration for their needs, but often end up with sub-optimal selections. In this paper, using molecular dynamics simulations as a case study, we propose a framework to guide users to select the optimal VM configurations that satisfy their requirements for scientific parallel computing in virtualized clusters. For molecular dynamics computation on a cluster of VMs, the guidance framework uses artificial neural networks which are trained to predict its execution times for various inputs, VM configurations, and status of interference among VMs. Using our performance prediction mechanisms, the guidance framework helps users choose an optimal or near-optimal VM cluster configuration under cost and runtime constraints.

An Online Auction Mechanism for Dynamic Virtual Cluster Provisioning in Geo-Distributed Clouds

It is common for cloud users to require clusters of inter-connected virtual machines (VMs) in a geo-distributed IaaS cloud, to run their services. Compared to isolated VMs, key challenges on dynamic virtual cluster (VC) provisioning (computation + communication resources) lie in two folds: (1) optimal placement of VCs and inter-VM traffic routing involve NP-hard problems, which are non-trivial to solve offline, not to mention if an online efficient algorithm is sought; (2) an efficient pricing mechanism is missing, which charges a market-driven price for each VC as a whole upon request, while maximizing system efficiency or provider revenue over the entire span. This paper proposes efficient online auction mechanisms to address the above challenges. We first design SWMOA, a novel online algorithm for dynamic VC provisioning and pricing, achieving truthfulness, individual rationality, computation efficiency, and $(1+2\log \mu)$ -competitiveness in social welfare, where $\mu$ is related to the problem size. Next, applying a randomized reduction technique, we convert the social welfare maximizing auction into a revenue maximizing online auction, PRMOA, achieving $O(\log \mu)$ -competitiveness in provider revenue, as well as truthfulness, individual rationality and computation efficiency. We investigate auction design in different cases of resource cost functions in the system. We validate the efficacy of the mechanisms through solid theoretical analysis and trace-driven simulations.

Deployment method of VM cluster based on graph theory for cloud resource management

Cloud computing is the next generation of computation and is regarded as the fifth utility service. A core issue of cloud computing is the deployment of virtual machine (VM). A new method had been constructed for VM cluster based on graph theory in this study. The authors first described VM cluster by energy minimisation. Then, they changed deployment of VM cluster into maximum flow minimum cut problem, added sink and source point, and found activity node. Finally, activity tail-nodes were connected and cut formed for VM cluster. Experimental results show that the segmentation method can effectively realise the VM deployment clusters, can reduce the overall bandwidth requirements after deployment.

Real-Time Earthquake Intensity Estimation Using Streaming Data Analysis of Social and Physical Sensors

Earthquake intensity is one of the key components of the decision-making process for disaster response and emergency services. Accurate and rapid intensity calculations can help to reduce total loss and the number of casualties after an earthquake. Modern intensity assessment procedures handle a variety of information sources, which can be divided into two main categories. The first type of data is that derived from physical sensors, such as seismographs and accelerometers, while the second type consists of data obtained from social sensors, such as witness observations of the consequences of the earthquake itself. Estimation approaches using additional data sources or that combine sources from both data types tend to increase intensity uncertainty due to human factors and inadequate procedures for temporal and spatial estimation, resulting in precision errors in both time and space. Here we present a processing approach for the real-time analysis of streams of data from both source types. The physical sensor data is acquired from the U.S. Geological Survey (USGS) seismic network in California and the social sensor data is based on Twitter user observations. First, empirical relationships between tweet rate and observed Modified Mercalli Intensity (MMI) are developed using data from the M6.0 South Napa, CAF earthquake that occurred on August 24, 2014. Second, the streams of both data types are analyzed together in simulated real-time to produce one intensity map. The second implementation is based on IBM InfoSphere Streams, a cloud platform for real-time analytics of big data. To handle large processing workloads for data from various sources, it is deployed and run on a cloud-based cluster of virtual machines. We compare the quality and evolution of intensity maps from different data sources over 10-min time intervals immediately following the earthquake. Results from the joint analysis shows that it provides more complete coverage, with better accuracy and higher resolution over a larger area than either data source alone.