双机容错系统中最佳检查点间隔的分析

设置检查点是容错计算机系统进行故障恢复的重要手段。因为检查点间隔选择过大或过小都将使系统性能受到影响，所以检查点间隔的适当选定是系统性能优化的一个重要指标。该文针对双机容错系统，采用检查点设置与回卷恢复的方法提出了一种系统模型，利用马尔科夫链得到了最佳检查点间隔的求解等式，通过实验证实了求解等式的正确性。     

超步诱导的回卷恢复

工作站机群系统已成为分布式并行处理发展的主流方向之一，随着机群系统应用领域的逐渐拓展和规模的不断扩大，人们对其可靠性的要求日益提高，设计高可靠的群机系统，需要着重研究其系统容错技术，本文叙述了并行异构回卷恢复和检查点派生，实现透明的可移植容错和负载均衡能力，避免调整检查点就构成全局一致性状态，不仅使BSP应用程序自治容错能力，而且能够在机群（Clusters）间迁移，保持系统负载均衡，重点介绍了检查点设置，检查点派生、卷回、进程迁移技术。     

基于虚拟化技术的仿真系统容错优化方法

节点崩溃或者仿真资源不足导致的分布式仿真系统故障,降低了仿真系统可靠性。为保证系统容错效果,降低容错开销,提出了一种基于虚拟化技术的仿真系统容错方法,按照系统故障发生的位置,对不同类型故障动态采用不同类型的容错策略。分析了检查点容错策略的优化方法,给出了最优设置间隔;结合虚拟化技术的优势,解决了副本容错策略的节点选择、副本数量以及位置分布问题;同时,引入基于虚拟机迁移的容错策略,并将其作为检查点容错策略和副本容错策略的补充,以降低容错开销。通过仿真实验数据对比,分析了动态容错策略与普通容错策略的性能,可知动态容错策略保证了系统容错性能,容错开销也保持在较低水平。     

用户指导的多层混合检查点技术及性能优化*

检查点机制是一种典型有效的软件容错技术。在对现有检查点实现技术综合研究的基础上,设计了一个用户指导的多层混合检查点模型uHybcr,并在IA64 Linux系统中予以实现。最后,通过对比测试对引入用户指导机制所带来的性能优化进行了验证。     

一种低费用的协调检查点算法

检查点算法作为一种有效的故障技术及容错手段,已广泛地运用在网格、分布式和云计算系统中。该文提出了一种非阻塞协调检查点算法,该算法增加了系统的可靠性,并允许检查点灵活设置,充分缩减了同步信息数量,加速了检查点形成时间。和典型的相关算法比较,该文提出的算法使用更少的同步控制消息,具有更低的费用,引入同步控制消息的时间复杂度由一般的O（n2）降到O（n）,且同步消息数仅仅为n-1。     

工作流系统适应性检查点机制的研究

容错技术已经成为工作流的研究热点,设置检查点是一种常用的容错方法。针对工作流系统提出一种适应性检查点机制,该机制通过最优化检查点数量和动态设置检查点间隔,大大提高了错误发生情况下任务按时完成的比率,并通过实验验证了该机制优于传统的检查点机制。     

基于内存缓存的异步检查点容错技术

高性能计算机系统规模越来越大,系统可靠性问题越来越严重.检查点技术是最典型的容错方法,但是因为并行文件系统的性能提高相对缓慢,数据写带宽低,传统检查点方法产生了严峻的性能问题.针对当前计算机系统计算和存储资源丰富,而并行文件系统写带宽提高相对滞后的特点,提出了基于内存缓存的异步检查点容错技术,传统的检查点技术被划分为两步:检查点文件首先被缓存在计算结点的局部内存,然后使用一个独立的帮助任务将数据拷贝到并行文件系统.利用局部内存带宽高以及帮助任务和计算任务并行执行的特点,新方法极大减小了检查点容错引入的时间开销,模拟和实际程序测试验证了异步检查点容错技术的有效性.     

协同系统中可扩展的动态容错模型研究及实现

针对协同系统的特点以及其对可靠性、实时性和性能等方面的要求，提出了一种可扩展的动态容错模型xDFT（Extensible Dynamic Fault Tolerance Model，xDFT）．本容错模型将容错支持结构和容错机制结合在一起：在本模型中首先定义了一种可扩展的、动态容错支撑结构；基于此，提出了一种容错机制．xDFT模型能够根据系统负载状况设定服务节点的负载阈值，改变服务冗余度，不仅提高了服务效率，而且以一种简单有效的方式实现了负载平衡．     

移动代理的容错研究

检查点是基于移动代理的分布系统容错的重要手段,讨论了检查点技术应用的两个关键问题,即何时以及如何建立检查点;利用马尔可夫模型计算检查点建立的最优时间和一个检查点建立算法。     

分层检查点的近似最优周期计算模型

针对大规模高性能计算（HPC）系统中检查点效率提升问题,提出一种面向分层检查点近似最优周期计算模型。首先,通过分析一个HPC系统中应用程序的执行过程,将检查点周期优化抽象为一个非线性的检查点成本模型;其次,通过分析可能故障位置推导出分层检查点成本公式,并引入两个减速因子和一个加速因子来模拟消息日志对分层检查点造成的影响。仿真实验结果表明,所提模型与理论近似最优周期检查点成本平均误差在5%以下,相对传统检查点周期优化模型的平均误差降低了20%,能够有效提高检查点的效率,提升HPC系统可用性。     

Finding a suitable checkpoint and recovery protocol for a distributed application

Checkpoint and recovery protocols are commonly used in distributed applications for providing fault tolerance. The performance of a checkpoint and recovery protocol is judged by the amount of computation it can save against the amount of overhead it incurs. This performance depends on different system and application characteristics, as well as protocol specific parameters. Hence, no single checkpoint and recovery protocol works equally well for all applications, and given a distributed application and a system it will run on, it is important to choose a protocol that will give the best performance for that system and application. In this paper, we present a scheme to automatically identify a suitable checkpoint and recovery protocol for a given distributed application running on a given system. The scheme involves a novel technique for finding the similarity between the communication pattern of two distributed applications that is of independent interest also. The similarity measure is based on a graph similarity problem. We present a heuristic for the graph similarity problem. Extensive experimental results are shown both for the graph similarity heuristic and the automatic identification scheme to show that an appropriate checkpoint and recovery protocol can be chosen automatically for a given application.     

Optimal checkpointing interval for two-level recovery schemes

It is important to design computer systems to tolerate some failures. This paper proposes two-level recovery schemes, soft checkpoint (SC) and hard checkpoint (HC), which are useful to recover from failures. Soft checkpoint is less reliable and less overhead than those of HC, and is set up between HCs to reduce the overhead of the process. The total expected overhead of one cycle from HC to HC is obtained, using Markov renewal processes, and an optimal interval which minimizes it is computed. It is shown in a numerical example that a two-level recovery scheme can achieve a good performance.     

The performance of cache-based error recovery in multiprocessors

Several variations of cache-based checkpointing for rollback error recovery from transient errors in shared-memory multiprocessors have been recently developed. By modifying the cache replacement policy, these techniques use the inherent redundancy in the memory hierarchy to periodically checkpoint the computation state. Three schemes, different in the manner in which they avoid rollback propagation, are evaluated in this paper. By simulation with address traces from parallel applications running on an Encore Multimax shared-memory multiprocessor, we evaluate the performance effect of integrating the recovery schemes in the cache coherence protocol. Our results indicate that the cache-based schemes can provide checkpointing capability with low performance overhead, but with uncontrollable high variability in the checkpoint interval     

一种降低并行程序检查点开销的方法

检查点设置和卷回恢复是提高系统可靠性和实现容错计算的有效途径,其性能通常用开销率来评价,而检查点开销是影响开销率的主要因素。针对目前并行程序运行时存在较多通信阻塞时间的现状,该文在写时复制检查点缓存的基础上提出了一种进一步降低检查点开销的方法。通过控制状态保存线程的调度和选择合适的状态保存粒度,该方法能很好地利用通信阻塞时间隐藏状态保存线程运行时带来的开销,从而能进一步降低开销率。     

Error recovery in shared memory multiprocessors using privatecaches

The problem of recovering from processor transient faults in shared memory multiprocessor systems is examined. A user-transparent checkpointing and recovery scheme using private caches is presented. Processes can recover from errors due to faulty processors by restarting from the checkpointed computation state. Implementation techniques using checkpoint identifiers and recovery stacks are examined as a means of reducing performance degradation in processor utilization during normal execution. This cache-based checkpointing technique prevents rollback propagation, provides rapid recovery, and can be integrated into standard cache coherence protocols. An analytical model is used to estimate the relative performance of the scheme during normal execution. Extensions to take error latency into account are presented     

Optimizing checkpoint-based fault-tolerance in distributed stream processing systems: Theory to practice

Fault-tolerance is an essential part of a stream processing system that guarantees data analysis could continue even after failures. State-of-the-art distributed stream processing systems use checkpointing to support fault-tolerance for stateful computations where the state of the computations is periodically persisted. However, the frequency of performing checkpoints impacts the performance (utilization, latency, and throughput) of the system as the checkpointing process consumes resources and time that can be used for actual computations. In practice, systems are often configured to perform checkpoints based on crude values ignoring factors such as checkpoint and restart costs, leading to suboptimal performance. In our previous work, we proposed a theoretical optimal checkpoint interval that maximizes the system utilization for stream processing systems to minimize the impact of checkpointing on system performance. In this article, we investigate the practical benefits of our proposed theoretical optimal by conducting experiments in a real-world cloud setting using different streaming applications; we use Apache Flink, a well-known stream processing system for our experiments. The experiment results demonstrate that an optimal interval can achieve better utilization, confirming the practicality of the theoretical model when applied to real-world applications. We observed utilization improvements from 10% to 200% for a range of failure rates from 0.3 failures per hour to 0.075 failures per minute. Moreover, we explore how performance measures: latency and throughput are affected by the optimal interval. Our observations demonstrate that significant improvements can be achieved using the optimal interval for both latency and throughput.     

面向更新密集型应用的内存数据库高效检查点技术

面向更新密集型应用的内存数据库系统,其检查点技术应符合几个关键的要求,包括检查点操作对正常事务处理的干扰尽可能小、能够处理存取倾斜状况、支持数据库系统的快速恢复、提供恢复过程中的系统可用性等.该文提出一种事务一致的分区检查点技术,采用基于元组的动态多版本并发控制机制,避免了读写事务的加锁冲突,提高系统吞吐能力;检查点操作以只读事务形式实现,存多版本并发控制下,避免检查点操作对正常事务处理的堵塞;由于检查点文件是事务一致的,只需要记录事务的Redo 日志信息,在系统恢复过程中,只需要对日志文件进行一遍扫描处理,加快恢复过程;基于优先级的数据分区装载和恢复,使得恢复过程中新事务的数据存取请求迅速得到满足,保证了恢复过程中的系统可用性.由于采用两级版本管理机制以及动态版本共享技术,多版本管理的空间开销降低到可以接受的水平.实验结果表明,文中提出的检查点技术方案获得比模糊检查点技术高27%的系统吞吐量,同时版本管理的空间开销在可接受的范围之内,满足高性能应用的要求.     

基于Lustre文件系统的MPI检查点系统实现技术与性能测试

基于协同式检查点的回卷恢复是在大规模并行计算机系统中得到采用的一项重要容错技术,其性能开销主要为协同协议和检查点映像存储所决定.描述了一个在MPICH2中实现的应用透明的并行检查点系统,相比已有的技术,该系统有以下特点：1）协同协议操作利用了并行应用的近邻通信特性,通过虚连接方法减少协议的处理开销;2）采用Lustre文件系统简化检查点映像文件管理的复杂性;3）通过并行I/O操作提高性能,优化检查点映像的存储过程.实际应用的测试表明,该检查点系统具有较小的运行时间开销和良好的可扩展性.     

数据流Eager传输:一种分布式流体系结构中的性能优化技术

分布式流体系结构扩展了分布式环境下的流计算模型,可在互联网上为大数据计算应用提供高效低成本的运行环境。互联网中较长的通信开销制约了计算性能。提出了一种数据流Eager传输的性能优化技术,以挖掘计算与通信之间的并行性,隐藏通信延迟。在分布式流体系结构原型系统中实现了该技术。实验结果表明,应用程序采用该优化技术之后的平均时间开销减少了19.58%,表明该优化技术能够显著提高应用的性能,具有良好的应用前景。