Добірка наукової літератури з теми "Virtualization, Scheduling, Heterogeneous Clusters, High Performance Computing"

Cloud computing provides on-demand computing and storage services with high performance and high scalability. However, the rising energy consumption of cloud data centers has become a prominent problem. In this paper, we first introduce an energy-aware framework for task scheduling in virtual clusters. The framework consists of a task resource requirements prediction module, an energy estimate module, and a scheduler with a task buffer. Secondly, based on this framework, we propose a virtual machine power efficiency-aware greedy scheduling algorithm (VPEGS). As a heuristic algorithm, VPEGS estimates task energy by considering factors including task resource demands, VM power efficiency, and server workload before scheduling tasks in a greedy manner. We simulated a heterogeneous VM cluster and conducted experiment to evaluate the effectiveness of VPEGS. Simulation results show that VPEGS effectively reduced total energy consumption by more than 20% without producing large scheduling overheads. With the similar heuristic ideology, it outperformed Min-Min and RASA with respect to energy saving by about 29% and 28%, respectively.

Efficient task scheduling of concurrent tasks is one of the primary requirements for high-performance computing platforms. Recent advances in high-performance computing have resulted in widespread performance improvement though at the cost of increased energy consumption and other system resources. In this article, an energy conscious scheduling algorithm with controlled threshold has been developed for precedence-constrained tasks on heterogeneous cluster, which aims at lower makespan along with reduced energy consumption. Energy conscious scheduling with controlled threshold algorithm combines the benefits of dynamic voltage scaling with controlled threshold-based duplication strategy to achieve its objectives. Effectiveness of the proposed algorithm is analyzed in comparison with available duplication- and non-duplication-based scheduling algorithms (with and without dynamic voltage scaling approach) to ascertain its performance and energy consumption. Exhaustive simulation results on random and real-world graphs demonstrate that energy conscious scheduling algorithm with controlled threshold has the potential to reduce energy consumption and makespan.

In this paper, we present a programming and runtime framework that enables the autonomic management of complex application workflows on hybrid computing infrastructures. The framework is designed to address system and application heterogeneity and dynamics to ensure that application objectives and constraints are satisfied. The need for such autonomic system and application management is becoming critical as computing infrastructures become increasingly heterogeneous, integrating different classes of resources from high-end HPC systems to commodity clusters and clouds. For example, the framework presented in this paper can be used to provision the appropriate mix of resources based on application requirements and constraints. The framework also monitors the system/application state and adapts the application and/or resources to respond to changing requirements or environment. To demonstrate the operation of the framework and to evaluate its ability, we employ a workflow used to characterize an oil reservoir executing on a hybrid infrastructure composed of TeraGrid nodes and Amazon EC2 instances of various types. Specifically, we show how different applications objectives such as acceleration, conservation and resilience can be effectively achieved while satisfying deadline and budget constraints, using an appropriate mix of dynamically provisioned resources. Our evaluations also demonstrate that public clouds can be used to complement and reinforce the scheduling and usage of traditional high performance computing infrastructure.

Cloud computing is Internet ("cloud") based development and use of computer technology ("computing"). It is a style of computing in which dynamically scalable and often virtualized resources are provided as a service over the Internet. This research deals with the balancing of work load in cloud environment. Load balancing is one of the essential factors to enhance the working performance of the cloud service provider. Grid computing utilizes the distributed heterogeneous resources in order to support complicated computing problems. Grid can be classified into two types: computing grid and data grid. We propose an improved load balancing algorithm for job scheduling in the Grid environment. Hence, in this research work, a multi-objective load balancing algorithm has been proposed to avoid deadlocks and to provide proper utilization of all the virtual machines (VMs) while processing the requests received from the users by VM classification. The capacity of virtual machine is computed based on multiple parameters like MIPS, RAM and bandwidth. Heterogeneous virtual machines of different MIPS and processing power in multiple data centers with different hosts have been created in cloud simulator. The VM’s are divided into 2 clusters using K-Means clustering mechanism in terms of processor MIPS, memory and bandwidth. The cloudlets are divided into two categories like High QOS and Low QOS based on the instruction size. The cloudlet whose task size is greater than the threshold value will enter into High QOS and cloudlet whose task size is lesser than the threshold value will enter into Low QOS. Submit the job of the user to the datacenter broker. The job of the user is submitted to the broker and it will first find the suitable VM according to the requirements of the cloudlet and will match VM depending upon its availability. Multiple parameters have been evaluated like waiting time, turnaround time, execution time and processing cost. This modified algorithm has an edge over the original approach in which each cloudlet build their own individual result set and it is later on built into a complete solution.

Abstract Cloud computing deals with voluminous heterogeneous data, and there is a need to effectively distribute the load across clusters of nodes to achieve optimal performance in terms of resource usage, throughput, response time, reliability, fault tolerance, and so on. The swarm intelligence methodologies use artificial intelligence to solve computationally challenging problems like load balancing, scheduling, and resource allocation at finite time intervals. In literature, sufficient works are being carried out to address load balancing problem in the cloud using traditional swarm intelligence techniques like ant colony optimization, particle swarm optimization, cuckoo search, bat optimization, and so on. But the traditional swarm intelligence techniques have issues with respect to convergence rate, arriving at the global optimum solution, complexity in implementation and scalability, which limits the applicability of such techniques in cloud domain. In this paper, we look into performance modeling aspects of some of the recent competitive swarm artificial intelligence based techniques like the whale, spider, dragonfly, and raven which are used for load balancing in the cloud. The results and analysis are presented over performance metrics such as total execution time, response time, resource utilization rate, and throughput achieved, and it is found that the performance of the raven roosting algorithm is high compared to other techniques.

We study the problem of scheduling applications composed of a large number of tasks on heterogeneous clusters. Tasks are identical, independent from each other, and can hence be computed in any order. The goal is to execute all the tasks as quickly as possible. We use the Master-Worker paradigm, where tasks are maintained by the master which will hand out batches of a variable amount of tasks to requesting workers. We introduce a new scheduling strategy, the Monitor strategy, and compare it to other strategies suggested in the literature. An image filtering application, known as matched filtering, has been used to compare the different strategies. Our implementation involves datastaging techniques in order to circumvent the possible bottleneck incurred by the master, and multi-threading to prevent possible processor idleness.

The computing needs and the input and result datasets of modern scientific and enterprise applications are growing exponentially. To support such applications, High-Performance Computing (HPC) systems need to employ thousands of cores and innovative data management. At the same time, an emerging trend in designing HPC systems is to leverage specialized asymmetric multicores, such as IBM Cell and AMD Fusion APUs, and commodity computational accelerators, such as programmable GPUs, which exhibit excellent price to performance ratio as well as the much needed high energy efficiency. While such accelerators have been studied in detail as stand-alone computational engines, integrating the accelerators into large-scale distributed systems with heterogeneous computing resources for data-intensive computing presents unique challenges and trade-offs. Traditional programming and resource management techniques cannot be directly applied to many-core accelerators in heterogeneous distributed settings, given the complex and custom instruction sets architectures, memory hierarchies and I/O characteristics of different accelerators. In this dissertation, we explore the design space of using commodity accelerators, specifically IBM Cell and programmable GPUs, in distributed settings for data-intensive computing and propose an adaptive framework for programming and managing heterogeneous clusters. The proposed framework provides a MapReduce-based extended programming model for heterogeneous clusters, which distributes tasks between asymmetric compute nodes by considering workload characteristics and capabilities of individual compute nodes. The framework provides efficient data prefetching techniques that leverage general-purpose cores to stage the input data in the private memories of the specialized cores. We also explore the use of an advanced layered-architecture based software engineering approach and provide mixin-layers based reusable software components to enable easy and quick deployment of heterogeneous clusters. The framework also provides multiple resource management and scheduling policies under different constraints, e.g., energy-aware and QoS-aware, to support executing concurrent applications on multi-tenant heterogeneous clusters. When applied to representative applications and benchmarks, our framework yields significantly improved performance in terms of programming efficiency and optimal resource management as compared to conventional, hand-tuned, approaches to program and manage accelerator-based heterogeneous clusters.
Ph. D.

A challenge posed by future computer architectures is the efficient exploitation of their many and sometimes heterogeneous computational cores. This challenge is exacerbated by the multiple facilities for data movement and sharing across cores resident on such platforms. To answer the question of how systems software should treat heterogeneous resources, this dissertation describes an approach that (1) creates a common manageable pool for all the resources present in the platform, and then (2) provides virtual machines (VMs) with multiple `personalities', flexibly mapped to and efficiently run on the heterogeneous underlying hardware. A VM's personality is its execution context on the different types of available processing resources usable by the VM. We provide mechanisms for making such platforms manageable and evaluate coordinated scheduling policies for mapping different VM personalities on heterogeneous hardware. Towards that end, this dissertation contributes technologies that include (1) restructuring hypervisor and system functions to create high performance environments that enable flexibility of execution and data sharing, (2) scheduling and other resource management infrastructure for supporting diverse application needs and heterogeneous platform characteristics, and (3) hypervisor level policies to permit efficient and coordinated resource usage and sharing. Experimental evaluations on multiple heterogeneous platforms, like one comprised of x86-based cores with attached NVIDIA accelerators and others with asymmetric elements on chip, demonstrate the utility of the approach and its ability to efficiently host diverse applications and resource management methods.

Cluster computing has recently gone through an evolution from single processor systems to multicore/multi-socket systems. This has resulted in lowering the cost/performance ratio of the compute machines. Compute farms that host these machines tend to become heterogeneous over time due to incremental extensions, hardware upgrades and/or nodes being purchased for users with particular needs. This heterogeneity is not surprising given the wide range of processor, memory and network technologies that become available and the relatively small price difference between these various options. Different CPU architectures, memory capacities, communication and I/O interfaces of the participating compute nodes present many challenges to job scheduling and often result in under or over utilization of the compute resources. In general, it is not feasible for the application programmers to specifically optimize their programs for such a set of differing compute n odes, due to the difficulty and time-intensiveness of such a task. The trend of heterogeneous compute farms has coincided with resurgence in the virtualization technology. Virtualization technology is receiving widespread adoption, mainly due to the benefits of server consolidation and isolation, load balancing, security and fault tolerance. Virtualization has also generated considerable interest in the High Performance Computing (HPC) community, due to the resulting high availability, fault tolerance, cluster partitioning and accommodation of conflicting user requirements. However, the HPC community is still wary of the potential overheads associated with‘ virtualization, as it results in slower network communications and disk I/O, which need to be addressed. The live migration feature, available to most virtualization technologies, can be leveraged to improve the throughput of a heterogeneous compute farm (HC) used for HPC applications. For this we mitigated the slow network communication in Xen; an open source virtual machine monitor. We present a detailed analysis of the communication framework of Xen and propose communication configurations that give 50% improvement over the conventional Xen network configuration. From a detailed study of the migration facility in Xen, we propose an improvement in the live migration facility specifically targeting HPC applications. This optimization gives around 50% improvement over the default migration facility of Xen. In this thesis, we also investigate resource scheduling in heterogeneous compute farm with the perspective of dynamic resource re-mapping. Our approach is to profile each job in the compute farm at runtime, and propose a better resource mapping compared to the initial allocation. We then migrate the job(s) to the best-suited homogeneous sub-cluster to improve overall throughput of the HC. For this, we develop a novel heterogeneity and virtualization-aware profiling framework, which is able to predict the CPU and communication characteristics of high performance scientific applications. The prediction accuracy of our performance estimation model is over 80%. The framework implementation is lightweight, with an overhead of 3%. Our experiments show that we are able to improve the throughput of the compute farm by 25% and the time saved by the HC with our framework is over 30%. The framework can be readily extended to HCs supporting a cloud computing environment.

Current advances in computer architectures have transformed clusters, traditional high performance distributed platforms, into hierarchical environments, where each node has multiple heterogeneous processing units including accelerators such as GPUs. Although parallel heterogeneous environments are becoming common, their efficient use is still an open problem. Current tools for development of parallel applications are mainly concerning with the exclusive use of accelerators, while it is argued that the adequate coordination of heterogeneous computing cores can significantly improve performance. The approach taken in this chapter to efficiently use such environments, which is experimented in the context of replicated dataflow applications, consists of scheduling processing tasks according to their characteristics and to the processors specificities. Thus, we can better utilize the available hardware as we try to execute each task into the best-suited processor for it. The proposed approach has been evaluated using two applications, for which there were previously available CPU-only and GPU-only implementations. The experimental results show that using both devices simultaneously can improve the performance significantly; moreover, the proposed method doubled the performance of a demand driven approach that utilizes both CPU and GPU, on the two applications in several scenarios.