基于GlusterFS的OpenStack平台设计

将GlusterFS应用于OpenStack中替代swift模块,并针对读写性能进行测试,实验证明在GlusterFS下虚拟机读写性能指标IOPS明显提高。     

基于共享存储池的元数据服务器机群的设计研究

在大型分布式对象存储系统中,元数据服务系统是一个潜在的访问瓶颈.本文提出一种通过分层式文件系统构建共享存储池的采用两次分布式哈希函数方式的元数据服务系统.其具有不需要人工干预的故障恢复性和易扩展特性,而且最大程度减少了MDS之间大量元数据的物理迁移.实验测试证明系统具有良好的I/O性能.     

集群文件系统的性能最优化配置

集群文件系统是当前存储系统的研究热点.在资源一定的条件下,存储系统中元数据服务器和数据存储服务器节点及客户端节点数量之间的配置比例会对系统性能产生较大的影响.分析了lustre集群文件系统的参数配置,针对两种典型应用环境:文件服务和Web服务,测试了不同节点数和不同lustre条块大小配置下的系统性能;通过对比分析,得出lustre集群文件系统最优化配置,为提高集群文件系统性能提供参考.测试结果显示.当lustre文件系统的OST节点数与client节点数相当时系统性能最好.     

DCFS2的元数据一致性策略

随着集群应用对机群文件系统的性能、容量和规模等需求的日益增长,采用多元数据服务器是机群文件系统发展的必然趋势．基于多元数据服务器的分布式元数据处理是文件系统研究的一个重要问题．机群文件系统DCFS2采用分布式日志技术和改进的两阶段提交协议解决了分布式元数据处理下元数据的一致性问题．性能测试结果表明,DCFS2所采用的基于分布式日志的元数据处理策略能够提供高的I／O性能,并能够保证在元数据服务器失效后文件系统快速恢复．     

一种适用于集群文件系统分级存储的分布式锁设计与分析

随着集群存储系统的规模越来越大,在一些系统搭建中出现分级存储的需求.目前集群文件系统的分布式锁管理器对于分级存储管理的数据迁移时,可能无法保证数据的一致性.参考传统分级存储系统的设计中,这里提出了集群文件系统布局锁的概念,通过布局锁来维护文件迁移时数据的一致性.通过分析目前的分布式锁机制在应用于集群文件系统分级功能时的不足,针对性地阐述了适用于集群文件系统分级存储的分布式布局锁,并对分布式布局锁的实现进行了分析.     

利用VMware在一台PC上实现xPC Target实时仿真

当前,实时仿真主要采用dSpace和xPC Target。前者价格昂贵,后者需要两台PC连接使用。由于现在PC性能的不断提高以及虚拟机技术的应用,可以仅在一台PC上实现实时的离线仿真。本文利用VMware建立虚拟的目标机,成功地在一台PC上实现xPC Target环境的实时离线仿真。     

利用VMware在一台PC上实现xPC Target实时仿真

当前,实时仿真主要采用dSpace和xPC Target。前者价格昂贵,后者需要两台PC连接使用。由于现在PC性能的不断提高以及虚拟机技术的应用,可以仅在一台PC上实现实时的离线仿真。本文利用VMware建立虚拟的目标机,成功地在一台PC上实现xPC Target环境的实时离线仿真。     

蓝鲸集群文件系统中资源交互一致性协议

在蓝鲸集群文件系统中,分布式资源交互在系统异常的情况下会出现资源状态不一致的情况,为解决这一问题,该文提出分布式资源交互一致性协议S2PC-RT。S2PC-RT引入资源交互序号保证一致性,通过增加资源申请空间减少消息等待。证明协议的正确性,并在蓝鲸集群文件系统中实现了协议。测试结果表明,S2PC-RT能够保证资源的一致性,有效提高分布式资源的交互性能。     

小规模集群文件系统中两级元数据服务器的设计与实现

在集群文件系统中,元数据服务器是整个系统正常运转的核心,它的可靠性和性能是设计系统时需要着重考虑的问题之一.本文设计了一个具有高可靠性、高性能的两级元数据服务器系统,兼顾了集中式元数据管理和分布式元数据管理的优点.系统中高级元数据服务器负责维护文件系统全局的目录结构和管理整个文件系统的命名空间,双元数据服务器负责维护文件元数据的分布信息,并采用了马尔可夫回报模型对两级元数据服务器系统进行了可靠性分析.实验数据表明,具有两级元数据服务器的集群文件系统能提供高吞吐量.     

MDDS:一种面向高性能计算的并行文件系统元数据性能提升方法

随着计算能力的增强、应用课题规模和复杂度的增加,高性能计算机对并行文件系统性能要求越来越高.在海量小文件和大规模并发I/O操作的应用场景中,文件系统元数据的吞吐率成为限制其性能的关键因素.设计并实现了元数据代理(meta data delegation service,MDDS),通过降低元数据服务间的耦合度,保证元数据集群的高可用性;使用目录子树方式管理元数据代理空间,避免跨节点目录引入的分布式原子操作的复杂性和低效性.并针对高性能计算中I/O转发架构,提出基于元数据代理的两种作业调度策略——单作业独占单元数据代理调度和多作业共享多元数据代理调度——实现作业间和作业内的负载均衡.在116台存储服务器上对MDDS进行评估,实验结果表明,元数据代理提供了拟线性的元数据性能,在大规模的环境中较Lustre CMD方案有较好的扩展性;两种调度方式有效分散了作业元数据的负载,改善了高性能计算中的元数据瓶颈问题.     

Overcoming data locality: An in-memory runtime file system with symmetrical data distribution

In many-task computing (MTC), applications such as scientific workflows or parameter sweeps communicate via intermediate files; application performance strongly depends on the file system in use. The state of the art uses runtime systems providing in-memory file storage that is designed for data locality: files are placed on those nodes that write or read them. With data locality, however, task distribution conflicts with data distribution, leading to application slowdown, and worse, to prohibitive storage imbalance. To overcome these limitations, we present MemFS, a fully symmetrical, in-memory runtime file system that stripes files across all compute nodes, based on a distributed hash function. Our cluster experiments with Montage and BLAST workflows, using up to 512 cores, show that MemFS has both better performance and better scalability than the state-of-the-art, locality-based file system, AMFS. Furthermore, our evaluation on a public commercial cloud validates our cluster results. On this platform MemFS shows excellent scalability up to 1024 cores and is able to saturate the 10G Ethernet bandwidth when running BLAST and Montage.     

Developing resource consolidation frameworks for moldable virtual machines in clouds

This paper considers the scenario where multiple clusters of Virtual Machines (i.e., termed Virtual Clusters) are hosted in a Cloud system consisting of a cluster of physical nodes. Multiple Virtual Clusters (VCs) cohabit in the physical cluster, with each VC offering a particular type of service for the incoming requests. In this context, VM consolidation, which strives to use a minimal number of nodes to accommodate all VMs in the system, plays an important role in saving resource consumption. Most existing consolidation methods proposed in the literature regard VMs as “rigid” during consolidation, i.e., VMs’ resource capacities remain unchanged. In VC environments, QoS is usually delivered by a VC as a single entity. Therefore, there is no reason why VMs’ resource capacity cannot be adjusted as long as the whole VC is still able to maintain the desired QoS. Treating VMs as “moldable” during consolidation may be able to further consolidate VMs into an even fewer number of nodes. This paper investigates this issue and develops a Genetic Algorithm (GA) to consolidate moldable VMs. The GA is able to evolve an optimized system state, which represents the VM-to-node mapping and the resource capacity allocated to each VM. After the new system state is calculated by the GA, the Cloud will transit from the current system state to the new one. The transition time represents overhead and should be minimized. In this paper, a cost model is formalized to capture the transition overhead, and a reconfiguration algorithm is developed to transit the Cloud to the optimized system state with low transition overhead. Experiments have been conducted to evaluate the performance of the GA and the reconfiguration algorithm.     

虚拟环境下Web服务动态负载均衡策略改进

为了提高Web服务集群的伸缩性和自动化能力,从虚拟化和负载均衡两方面研究集群系统,对现有负载采集策略做了改进,设计并实现了一种可根据负载值自动控制集群规模的模型XCluster。新模型运行在Xen提供的虚拟化环境中,实时监视宿主机层和虚拟机层的负载状态,随着集群系统总负载的增长,逐渐引入新的虚拟机来扩大集群规模,同时将任务合理分配到各个虚拟机节点上;当总负载下降时,逐渐关闭虚拟机缩小集群规模,释放出来的硬件资源又可以提供给其他集群系统使用。理论分析和实验结果表明,XCluster只需占用很少的网络通信量完成信息收集和命令下达,能够充分利用虚拟机易于管理的优势完成后端节点的调度,并且在任务总量相同的情况下,使用尽可能少的集群节点来执行任务。     

WOW: Self-organizing Wide Area Overlay Networks of Virtual Workstations

This paper describes WOW, a distributed system that combines virtual machine, overlay networking and peer-to-peer techniques to create scalable wide-area networks of virtual workstations for high-throughput computing. The system is architected to: facilitate the addition of nodes to a pool of resources through the use of system virtual machines (VMs) and self-organizing virtual network links; to maintain IP connectivity even if VMs migrate across network domains; and to present to end-users and applications an environment that is functionally identical to a local-area network or cluster of workstations. We describe IPOP, a network virtualization technique that builds upon a novel, extensible user-level decentralized technique to discover, establish and maintain overlay links to tunnel IP packets over different transports (including UDP and TCP) and across firewalls. We evaluate latency and bandwidth overheads of IPOP and also time taken for a new node to become fully-routable over the virtual network. We also report on several experiments conducted on a testbed WOW deployment with 118 P2P router nodes over PlanetLab and 33 VMware-based VM nodes distributed across six firewalled domains. Experiments show that the testbed delivers good performance for two unmodified, representative benchmarks drawn from the life-sciences domain. We also demonstrate that the system is capable of seamlessly maintaining connectivity at the virtual IP layer for typical client/server applications (NFS, SSH, PBS) when VMs migrate across a WAN.     

虚拟化环境下面向多目标优化的自适应SSD缓存系统

以SSD为代表的新型存储介质在虚拟化环境下得到了广泛的应用,通常作为虚拟机读写缓存,起到优化磁盘IO性能的作用.已有研究往往关注SSD缓存的容量规划,依据缓存读写命中率评价SSD缓存分配效果,未能充分考虑SSD的服务能力上限,难以适用于典型的分布式应用场景,存在虚拟机抢占SSD缓存资源,导致虚拟机中应用性能违约的可能.本文实现了虚拟化环境下面向多目标优化的自适应SSD缓存系统,考虑了SSD的服务能力上限.基于自适应闭环实现对虚拟机和应用状态的动态感知.动态检测局部SSD缓存抢占状态,基于聚类方法生成虚拟机的优化放置方案,依据全局SSD缓存供给能力确定虚拟机迁移顺序和时机.实验结果表明该方法在应对典型分布式应用场景时可以有效缓解SSD缓存资源的争用,同时满足应用对虚拟机放置的需求,提升应用的性能并兼顾应用的可靠性.在Hadoop应用场景下,平均降低了25%的任务执行时间,对IO密集型应用平均提升39%的吞吐率.在ZooKeeper应用场景下,以不到5%的性能损失为代价应对了虚拟化主机的单点失效带来的虚拟机宕机问题.     

HPC Cloud环境中基于网络Ⅰ/O负载的虚拟机放置算法

随着虚拟化技术和云计算技术的发展,越来越多的高性能计算应用运行在云计算资源上.在基于虚拟化技术的高性能计算云系统中,高性能计算应用运行在多个虚拟机之中,这些虚拟机可能放置在不同的物理节点上.若多个通信密集型作业的虚拟机放置在相同的物理节点上,虚拟机之间将竞争物理节点的网络Ⅰ/O资源,如果虚拟机对网络Ⅰ/O资源的需求超过物理节点的网络Ⅰ/O带宽上限,将严重影响通信密集型作业的计算性能.针对虚拟机对网络Ⅰ/O资源的竞争问题,提出一种基于网络Ⅰ/O负载均衡的虚拟机放置算法NLPA,该算法采用网络Ⅰ/O负载均衡策略来减少虚拟机对网络Ⅰ/O资源的竞争.实验表明,与贪心算法进行比较,对于同样的高性能计算作业测试集,NLPA算法在完成作业的计算时间、系统中的网络Ⅰ/O负载吞吐率、网络Ⅰ/O负载均衡3个方面均有更好的表现.     

Virtual machine cluster mobility in inter-cloud platforms

Modern cloud computing applications developed from different interoperable services that are interfacing with each other in a loose coupling approach. This work proposes the concept of the Virtual Machine (VM) cluster migration, meaning that services could be migrated to various clouds based on different constraints such as computational resources and better economical offerings. Since cloud services are instantiated as VMs, an application can be seen as a cluster of VMs that integrate its functionality. We focus on the VM cluster migration by exploring a more sophisticated method with regards to VM network configurations. In particular, networks are hard to managed because their internal setup is changed after a migration, and this is related with the configuration parameters during the re-instantiation to the new cloud platform. To address such issue, we introduce a Software Defined Networking (SDN) service that breaks the problem of network configuration into tractable pieces and involves virtual bridges instead of references to static endpoints. The architecture is modular, it is based on the SDN OpenFlow protocol and allows VMs to be paired in cluster groups that communicate with each other independently of the cloud platform that are deployed. The experimental analysis demonstrates migrations of VM clusters and provides a detailed discussion of service performance for different cases.     

CEFT: A cost-effective,fault-tolerant parallel virtual file system

The vulnerability of computer nodes due to component failures is a critical issue for cluster-based file systems. This paper studies the development and deployment of mirroring in cluster-based parallel virtual file systems to provide fault tolerance and analyzes the tradeoffs between the performance and the reliability in the mirroring scheme. It presents the design and implementation of CEFT, a scalable RAID-10 style file system based on PVFS, and proposes four novel mirroring protocols depending on whether the mirroring operations are server-driven or client-driven, whether they are asynchronous or synchronous. The comparisons of their write performances, measured in a real cluster, and their reliability and availability, obtained through analytical modeling, show that these protocols strike different tradeoffs between the reliability and performance. Protocols with higher peak write performance are less reliable than those with lower peak write performance, and vice versa. A hybrid protocol is proposed to optimize this tradeoff.     

ROARS: a robust object archival system for data intensive scientific computing

As scientific research becomes more data intensive, there is an increasing need for scalable, reliable, and high performance storage systems. Such data repositories must provide both data archival services and rich metadata, and cleanly integrate with large scale computing resources. ROARS is a hybrid approach to distributed storage that provides both large, robust, scalable storage and efficient rich metadata queries for scientific applications. In this paper, we present the design and implementation of ROARS, focusing primarily on the challenge of maintaining data integrity across long time scales. We evaluate the performance of ROARS on a storage cluster, comparing to the Hadoop distributed file system and a centralized file server. We observe that ROARS has read and write performance that scales with the number of storage nodes, and integrity checking that scales with the size of the largest node. We demonstrate the ability of ROARS to function correctly through multiple system failures and reconfigurations. ROARS has been in production use for over three years as the primary data repository for a biometrics research lab at the University of Notre Dame.     

Performance models of storage contention in cloud environments

We propose simple models to predict the performance degradation of disk requests due to storage device contention in consolidated virtualized environments. Model parameters can be deduced from measurements obtained inside Virtual Machines (VMs) from a system where a single VM accesses a remote storage server. The parameterized model can then be used to predict the effect of storage contention when multiple VMs are consolidated on the same server. We first propose a trace-driven approach that evaluates a queueing network with fair share scheduling using simulation. The model parameters consider Virtual Machine Monitor level disk access optimizations and rely on a calibration technique. We further present a measurement-based approach that allows a distinct characterization of read/write performance attributes. In particular, we define simple linear prediction models for I/O request mean response times, throughputs and read/write mixes, as well as a simulation model for predicting response time distributions. We found our models to be effective in predicting such quantities across a range of synthetic and emulated application workloads.