Fault tolerance is a key challenge as high performance computing systems continue to increase component counts, individual component reliability decreases, and hardware and software complexity increases. To better understand the potential impacts of failures on next-generation systems, significant effort has been devoted to collecting, characterizing and analyzing failures on current systems. These studies require large volumes of data and complex analysis in an attempt to identify statistical properties of the failure data. In this paper, we examine the lifetime of failures on the Cielo supercomputer that was located at Los Alamos National Laboratory, looking specifically at the time between faults on this system. Through this analysis, we show that the time between uncorrectable faults for this system obeys Benford’s law, This law applies to a number of naturally occurring collections of numbers and states that the leading digit is more likely to be small, for example a leading digit of 1 is more likely than 9. We also show that a number of common distributions used to model failures also follow this law. This work provides critical analysis on the distribution of times between failures for extreme-scale systems. Specifically, the analysis in this work could be used as a simple form of failure prediction or used for modeling realistic failures.
The Message Passing Interface (MPI) remains the dominant programming model for scientific applications running on today's high-performance computing (HPC) systems. This dominance stems from MPI's powerful semantics for inter-process communication that has enabled scientists to write applications for simulating important physical phenomena. MPI does not, however, specify how messages and synchronization should be carried out. Those details are typically dependent on low-level architecture details and the message characteristics of the application. Therefore, analyzing an application's MPI resource usage is critical to tuning MPI's performance on a particular platform. The result of this analysis is typically a discussion of the mean message sizes, queue search lengths and message arrival times for a workload or set of workloads. While a discussion of the arithmetic mean in MPI resource usage might be the most intuitive summary statistic, it is not always the most accurate in terms of representing the underlying data. In this paper, we analyze MPI resource usage for a number of key MPI workloads using an existing MPI trace collector and discrete-event simulator. Our analysis demonstrates that the average, while easy and efficient to calculate, is a useful metric for characterizing latency and bandwidth measurements, but may not be a good representation of application message sizes, match list search depths, or MPI inter-operation times. Additionally, we show that the median and mode are superior choices in many cases. We also observe that the arithmetic mean is not the best representation of central tendency for data that are drawn from distributions that are multi-modal or have heavy tails. The results and analysis of our work provide valuable guidance on how we, as a community, should discuss and analyze MPI resource usage data for scientific applications.
Fault tolerance poses a major challenge for future large-scale systems. Current research on fault tolerance has been principally focused on mitigating the impact of uncorrectable errors: errors that corrupt the state of the machine and require a restart from a known good state. However, correctable errors occur much more frequently than uncorrectable errors and may be even more common on future systems. Although an application can safely continue to execute when correctable errors occur, recovery from a correctable error requires the error to be corrected and, in most cases, information about its occurrence to be logged. The potential performance impact of these recovery activities has not been extensively studied in HPC. In this paper, we use simulation to examine the relationship between recovery from correctable errors and application performance for several important extreme-scale workloads. Our paper contains what is, to the best of our knowledge, the first detailed analysis of the impact of correctable errors on application performance. Our study shows that correctable errors can have significant impact on application performance for future systems. We also find that although the focus on correctable errors is focused on reducing failure rates, reducing the time required to log individual errors may have a greater impact on overheads at scale. Finally, this study outlines the error frequency and durations targets to keep correctable overheads similar to that of today's systems. This paper provides critical analysis and insight into the overheads of correctable errors and provides practical advice to systems administrators and hardware designers in an effort to fine-tune performance to application and system characteristics.
The Message Passing Interface (MPI) remains the dominant programming model for scientific applications running on today's high-performance computing (HPC) systems. This dominance stems from MPI's powerful semantics for inter-process communication that has enabled scientists to write applications for simulating important physical phenomena. MPI does not, however, specify how messages and synchronization should be carried out. Those details are typically dependent on low-level architecture details and the message characteristics of the application. Therefore, analyzing an applications MPI usage is critical to tuning MPI's performance on a particular platform. The results of this analysis is typically a discussion of average message sizes for a workload or set of workloads. While a discussion of the message average might be the most intuitive summary statistic, it might not be the most useful in terms of representing the entire message size dataset for an application. Using a previously developed MPI trace collector, we analyze the MPI message traces for a number of key MPI workloads. Through this analysis, we demonstrate that the average, while easy and efficient to calculate, may not be a good representation of all subsets of application messages sizes, with median and mode of message sizes being a superior choice in most cases. We show that the problem with using the average relate to the multi-modal nature of the distribution of point-to-point messages. Finally, we show that while scaling a workload has little discernible impact on which measures of central tendency are representative of the underlying data, different input descriptions can significantly impact which metric is most effective. The results and analysis in this paper have the potential for providing valuable guidance on how we as a community should discuss and analyze MPI message data for scientific applications.
This paper explores key differences of MPI match lists for several important United States Department of Energy (DOE) applications and proxy applications. This understanding is critical in determining the most promising hardware matching design for any given high-speed network. The results of MPI match list studies for the major open-source MPI implementations, MPICH and Open MPI, are presented, and we modify an MPI simulator, LogGOPSim, to provide match list statistics. These results are discussed in the context of several different potential design approaches to MPI matching–capable hardware. The data illustrate the requirements for different hardware designs in terms of performance and memory capacity. This paper's contributions are the collection and analysis of data to help inform hardware designers of common MPI requirements and highlight the difficulties in determining these requirements by only examining a single MPI implementation.
Coordinated checkpoint/restart is currently the dominant approach to mitigating the impact of failures on important scientific applications running on large-scale distributed systems. However, there is widespread evidence that coordinated checkpointing may no longer be viable on next-generation systems. Uncoordinated checkpoint/restart attempts to address the shortcomings of coordinated checkpoint/restart by allowing application processes to checkpoint their state independently. However, eliminating coordination may significantly degrade application performance. In this paper, we propose an approach that leverages existing coordination in important scientific applications to approximately coordinate checkpoints. Specifically, we propose to extend MPI implementations to force checkpoints to occur immediately after the completion of a collective operation. We evaluate the performance implications of this approach using an existing validated simulation framework. Our results demonstrate that approximately coordinated checkpointing can significantly improve application performance relative to totally uncoordinated checkpointing. We also show that forcing checkpoints to occur following a collective operation has a small impact on the nominal checkpoint interval for several important workloads. As a whole, the results presented in this paper demonstrate that approximately coordinated checkpointing may provide significant performance benefits without significantly increasing the cost of failure recovery.
Concern about memory errors has been widespread in high-performance computing (HPC) for decades. These concerns have led to significant research on detecting and correcting memory errors to improve performance and provide strong guarantees about the correctness of the memory contents of scientific simulations. However, power concerns and changes in memory architectures threaten the viability of current approaches to protecting memory (e.g., Chipkill). Returning to less protective error-correcting codes (ECC), e.g., single-error correction, double-error detection (SECDED), may increase the frequency of memory errors, including silent data corruption (SDC). SDC has the potential to silently cause applications to produce incorrect results and mislead domain scientists. We propose an approach for exploiting unnecessary bits in pointer values to support encoding the pointer with a Reed-Solomon code. Encoding the pointer allows us to provides strong capabilities for correcting and detecting corruption of pointer values. In this paper, we provide a detailed description of how we can exploit unnecessary pointer bits to store Reed-Solomon parity symbols. We evaluate the performance impacts of this approach and examine the effectiveness of the approach against corruption. Our results demonstrate that encoding and decoding is fast (less than 45 per event) and that the protection it provides is robust (the rate of miscorrection is less than 5% even for significant corruption). The data and analysis presented in this paper demonstrates the power of our approach. It is fast, tunable, requires no additional per-pointer storage resources, and provides robust protection against pointer corruption.
Although its demise has been frequently predicted, the Message Passing Interface (MPI) remains the dominant programming model for scientific applications running on high-performance computing (HPC) systems. MPI specifies powerful semantics for interprocess communication that have enabled scientists to write applications for simulating important physical phenomena. However, these semantics have also presented several significant challenges. For example, the existence of wildcard values has made the efficient enforcement of MPI message matching semantics challenging. Significant research has been dedicated to accelerating MPI message matching. One common approach has been to offload matching to dedicated hardware. One of the challenges that hardware designers have faced is knowing how to size hardware structures to accommodate outstanding match requests. Applications that exceed the capacity of specialized hardware typically must fall back to storing match requests in bulk memory, e.g. DRAM on the host processor. In this paper, we examine the implications of hardware matching and develop guidance on sizing hardware matching structure to strike a balance between minimizing expensive dedicated hardware resources and overall matching performance. By examining the message matching behavior of several important HPC workloads, we show that when specialized hardware matching is not dramatically faster than matching in memory the offload hardware's match queue capacity can be reduced without significantly increasing match time. On the other hand, effectively exploiting the benefits of very fast specialized matching hardware requires sufficient storage resources to ensure that every search completes in the specialized hardware. The data and analysis in this paper provide important guidance for designers of MPI message matching hardware.
Attaining high performance with MPI applications requires efficient message matching to minimize message processing overheads and the latency these overheads introduce into application communication. In this paper, we use a validated simulation-based approach to examine the relationship between MPI message matching performance and application time-to-solution. Specifically, we examine how the performance of several important HPC workloads is affected by the time required for matching. Our analysis yields several important contributions: (i) the performance of current workloads is unlikely to be significantly affected by MPI matching unless match queue operations get much slower or match queues get much longer; (ii) match queue designs that provide sublinear performance as a function of queue length are unlikely to yield much benefit unless match queue lengths increase dramatically; and (iii) we provide guidance on how long the mean time per match attempt may be without significantly affecting application performance. The results and analysis in this paper provide valuable guidance on the design and development of MPI message match queues.
Proceedings - International Conference for High Performance Computing, Networking, Storage, and Analysis, SC 2018
Levy, Scott; Ferreira, Kurt B.; DeBardeleben, Nathan; Siddiqua, Taniya; Sridharan, Vilas; Baseman, Elisabeth
Maintaining the performance of high-performance computing (HPC) applications as failures increase is a major challenge for next-generation extreme-scale systems. Recent work demonstrates that hardware failures are expected to become more common. Few existing studies, however, have examined failures in the context of the entire lifetime of a single platform. In this paper, we analyze a corpus of empirical failure data collected over the entire five-year lifetime of Cielo, a leadership-class HPC system. Our analysis reveals several important findings about failures on Cielo: (i) its memory (DRAM and SRAM) exhibited no aging effects; detectable, uncorrectable errors (DUE) showed no discernible increase over its five-year lifetime; (ii) contrary to popular belief, correctable DRAM faults are not predictive of future uncorrectable DRAM faults; (iii) the majority of system down events have no identifiable hardware root cause, highlighting the need for more comprehensive logging facilities to improve failure analysis on future systems; and (iv) continued advances will be needed in order for current failure mitigation techniques to be viable on future systems. Our analysis of this corpus of empirical data provides critical analysis of, and guidance for, the deployment of extreme-scale systems.
2018 IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2018
Baseman, Elisabeth; Debardeleben, Nathan; Blanchard, Sean; Moore, Juston; Tkachenko, Olena; Ferreira, Kurt B.; Siddiqua, Taniya; Sridharan, Vilas
As the scale of high performance computing facilities approaches the exascale era, gaining a detailed understanding of hardware failures becomes important. In particular, the extreme memory capacity of modern supercomputers means that data corruption errors which were statistically negligible at smaller scales will become more prevalent. In order to understand hardware faults and mitigate their adverse effects on exascale workloads, we must learn from the behavior of current hardware. In this work, we investigate the predictability of DRAM errors using field data from two recently decommissioned supercomputers: Cielo, at Los Alamos National Laboratory, and Hopper, at Lawrence Berkeley National Laboratory. Due to the volume and complexity of the field data, we apply statistical machine learning to predict the probability of DRAM errors at previously un-accessed locations. We compare the predictive performance of six machine learning algorithms, and find that a model incorporating physical knowledge of DRAM spatial structure outperforms purely statistical methods. Our findings both support expected physical behavior of DRAM hardware as well as providing a mechanism for real-time error prediction. We demonstrate real-world feasibility by training an error model on one supercomputer and effectively predicting errors on another. Our methods demonstrate the importance of spatial locality over temporal locality in DRAM errors, and show that relatively simple statistical models are effective at predicting future errors based on historical data, allowing proactive error mitigation.
Input/output (I/O) from various sources often contend for scarcely available bandwidth. For example, checkpoint/restart (CR) protocols can help to ensure application progress in failure-prone environments. However, CR I/O alongside an application's normal, requisite I/O can increase I/O contention and might negatively impact performance. In this work, we consider different aspects (system-level scheduling policies and hardware) that optimize the overall performance of concurrently executing CR-based applications that share I/O resources. We provide a theoretical model and derive a set of necessary constraints to minimize the global waste on a given platform. Our results demonstrate that Young/Daly's optimal checkpoint interval, despite providing a sensible metric for a single, undisturbed application, is not sufficient to optimally address resource contention at scale. We show that by combining optimal checkpointing periods with contention-aware system-level I/O scheduling strategies, we can significantly improve overall application performance and maximize the platform throughput. Finally, we evaluate how specialized hardware, namely burst buffers, may help to mitigate the I/O contention problem. Altogether, these results provide critical analysis and direct guidance on how to design efficient, CR ready, large -scale platforms without a large investment in the I/O subsystem.
Attaining high performance with MPI applications requires efficient message matching to minimize message processing overheads and the latency these overheads introduce into application communication. In this paper, we use a validated simulation-based approach to examine the relationship between MPI message matching performance and application time-to-solution. Specifically, we examine how the performance of several important HPC workloads is affected by the time required for matching. Our analysis yields several important contributions: (i) the performance of current workloads is unlikely to be significantly affected by MPI matching unless match queue operations get much slower or match queues get much longer; (ii) match queue designs that provide sublinear performance as a function of queue length are unlikely to yield much benefit unless match queue lengths increase dramatically; and (iii) we provide guidance on how long the mean time per match attempt may be without significantly affecting application performance. The results and analysis in this paper provide valuable guidance on the design and development of MPI message match queues.
High-performance computing (HPC) systems are critically important to the objectives of universities, national laboratories, and commercial companies. Because of the cost of deploying and maintaining these systems ensuring their efficient use is imperative. Job scheduling and resource management are critically important to the efficient use of HPC systems. As a result, significant research has been conducted on how to effectively schedule user jobs on HPC systems. Developing and evaluating job scheduling algorithms, however, requires a detailed understanding of how users request resources on HPC systems. In this paper, we examine a corpus of job data that was collected on Trinity, a leadership-class supercomputer. During the stabilization period of its Intel Xeon Phi (Knights Landing) partition, it was made available to users outside of a classified environment for the Trinity Open Science Phase 2 campaign. We collected information from the resource manager about each user job that was run during this Open Science period. In this paper, we examine the jobs contained in this dataset. Our analysis reveals several important characteristics of the jobs submitted during the Open Science period and provides critical insight into the use of one of the most powerful supercomputers in existence. Specifically, these data provide important guidance for the design, development, and evaluation of job scheduling and resource management algorithms.
Proceedings - 2018 IEEE 32nd International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2018
Herault, Thomas; Robert, Yves; Bouteiller, Aurelien; Arnold, Dorian; Ferreira, Kurt B.; Bosilca, George; Dongarra, Jack
In high-performance computing environments, input/output (I/O) from various sources often contend for scarce available bandwidth. Adding to the I/O operations inherent to the failure-free execution of an application, I/O from checkpoint/restart (CR) operations (used to ensure progress in the presence of failures) place an additional burden as it increase I/O contention, leading to degraded performance. In this work, we consider a cooperative scheduling policy that optimizes the overall performance of concurrently executing CR-based applications which share valuable I/O resources. First, we provide a theoretical model and then derive a set of necessary constraints needed to minimize the global waste on the platform. Our results demonstrate that the optimal checkpoint interval, as defined by Young/Daly, despite providing a sensible metric for a single application, is not sufficient to optimally address resource contention at the platform scale. We therefore show that combining optimal checkpointing periods with I/O scheduling strategies can provide a significant improvement on the overall application performance, thereby maximizing platform throughput. Overall, these results provide critical analysis and direct guidance on checkpointing large-scale workloads in the presence of competing I/O while minimizing the impact on application performance.
Baseman, Elisabeth B.; DeBardeleben, Nathan D.; Blanchard, Sean B.; Moore, Juston M.; Tkachenko, Olena T.; Ferreira, Kurt B.; Siddiqua, Taniya S.; Sridharan, Vilas S.
With the increased scale expected on future leadership-class systems, detailed information about the resource usage and performance of MPI message matching provides important insights into how to maintain application performance on next-generation systems. However, obtaining MPI message matching performance data is often not possible without significant effort. A common approach is to instrument an MPI implementation to collect relevant statistics. While this approach can provide important data, collecting matching data at runtime perturbs the application’s execution, including its matching performance, and is highly dependent on the MPI library’s matchlist implementation. In this paper, we introduce a trace-based simulation approach to obtain detailed MPI message matching performance data for MPI applications without perturbing their execution. Using a number of key parallel workloads, we demonstrate that this simulator approach can rapidly and accurately characterize matching behavior. Specifically, we use our simulator to collect several important statistics about the operation of the MPI posted and unexpected queues. For example, we present data about search lengths and the duration that messages spend in the queues waiting to be matched. Data gathered using this simulation-based approach have significant potential to aid hardware designers in determining resource allocation for MPI matching functions and provide application and middleware developers with insight into the scalability issues associated with MPI message matching.
Aggregating millions of hardware components to construct an exascale computing platform will pose significant resilience challenges. In addition to slowdowns associated with detected errors, silent errors are likely to further degrade application performance. Moreover, silent data corruption (SDC) has the potential to undermine the integrity of the results produced by important scientific applications.In this paper, we propose an application-independent mechanism to efficiently detect and correct SDC in read-mostly memory, where SDC may be most likely to occur. We use memory protection mechanisms to maintain compressed backups of application memory. We detect SDC by identifying changes in memory contents that occur without explicit write operations. We demonstrate that, for several applications, our approach can potentially protect a significant fraction of application memory pages from SDC with modest overheads. Moreover, our proposed technique can be straightforwardly combined with many other approaches to provide a significant bulwark against SDC.
Current practice for mitigating DRAM hardwarefaults is to simply discard the entire faulty DIMM. However, this becomes increasingly expensive and wasteful as the priceof memory hardware increases and moves physically closer toprocessing units. Accurately characterizing memory faults inreal-time in order to pre-empt future potentially catastrophicfailures is crucial to conserving resources by blacklisting smallaffected regions of memory rather than discarding an entirehardware component. We further evaluate and extend a machinelearning method for DRAM fault characterization introduced inprior work by Baseman et al. at Los Alamos National Laboratory. We report on the usefulness of a variety of training sets, usinga set of production-relevant metrics to evaluate the method ondata from a leadership-class supercomputing facility. We observean increase in percent of faults successfully mitigated as well asa decrease in percent of wasted blacklisted pages, regardless oftraining set, when using the learned algorithm as compared to ahuman-expert, deterministic, and rule-based approach.
2017 IEEE Int. Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2017
Siddiqua, Taniya; Sridharan, Vilas; Raasch, Steven E.; Debardeleben, Nathan; Ferreira, Kurt B.; Levy, Scott; Baseman, Elisabeth; Guan, Qiang
In order to provide high system resilience, it is important to understand the nature of the faults that occur in the field. This study analyzes fault rates from a production system that has been monitored for five years, capturing data for the entire operational lifetime of the system. The data show that devices in this system did not show any sign of aging during the monitoring period, suggesting that the lifetime of a system may be longer than five years. In DRAM, the relative incidence of fault modes changed insignificantly over the system's lifetime: The relative rate of each fault mode at the end of the system's lifetime was within 1.4 percentage point of the rate observed during the first year. SRAM caches in the system exhibited different fault modes including cache-way fault and single-bit faults. Overall, this study provides insights on how fault modes and types in a system evolve over the system's lifetime.
Fault-tolerance poses a major challenge for future large-scale systems. Active research into coordinated, uncoordinated, and hybrid checkpointing systems has explored how the introduction of asynchrony can address anticipated scalability issues. While fully uncoordinated approaches have been shown to have significant delays, the degree of sychronization required to keep overheads low has not yet been significantly addressed. In this paper, we use a simulation-based approach to show the impact of synchronization on local checkpoint activity. Specifically, we show the degree of synchronization needed to keep the impacts of local checkpointing low is attainable with current technology for a number of key production HPC workloads. Our work provides a critical analysis and comparison of synchronization and local checkpointing. This enables users and system administrators to fine-tune the checkpointing scheme to the application and system characteristics available.
Scientific workloads running on current extreme-scale systems routinely generate tremendous volumes of data for postprocessing. This data movement has become a serious issue due to its energy cost and the fact that I/O bandwidths have not kept pace with data generation rates. In situ analytics is an increasingly popular alternative in which post-simulation processing is embedded into an application, running as part of the same MPI job. This can reduce data movement costs but introduces a new potential source of interference for the application. Using a validated simulation-based approach, we investigate how best to mitigate the interference from time-shared in situ tasks for a number of key extreme-scale workloads. This paper makes a number of contributions. First, we show that the independent scheduling of in situ analytics tasks can significantly degradation application performance, with slowdowns exceeding 1000%. Second, we demonstrate that the degree of synchronization found in many modern collective algorithms is sufficient to significantly reduce the overheads of this interference to less than 10% in most cases. Finally, we show that many applications already frequently invoke collective operations that use these synchronizing MPI algorithms. Therefore, the syncronization introduced by these MPI collective algorithms can be leveraged to efficiently schedule analytics tasks with minimal changes to existing applications. This paper provides critical analysis and guidance for MPI users and developers on the importance of scheduling in situ analytics tasks. It shows the degree of synchronization needed to mitigate the performance impacts of these time-shared coupled codes and demonstrates how that synchronization can be realized in an extreme-scale environment using modern collective algorithms.
As high-performance computing systems continue to grow in scale and complexity, the study of faults and errors is critical to the design of future systems and mitigation schemes. Fault modes in system DRAM are a frequently-investigated key aspect of memory reliability. While current schemes require offline analysis for proper classification, current state-of-the-art mitigation techniques require accurate online prediction for optimal performance. In this work, we explore the predictive performance of an online machine learning-based approach in classifying DRAM fault modes from two leadership-class supercomputing facilities. Our results compare the predictive performance of this online approach with the current rule-based approach based on expert knowledge, finding a 12% predictive performance improvement. We also investigate the universality of our classifiers by evaluating predictive performance using training data from disparate computing systems to achieve a 7% improvement in predictive performance. Our work provides a critical analysis of this online learning technique and can benefit system designers to help inform best practices for dealing with reliability on future systems.
Next-generation applications increasingly rely on in situ analytics to guide computation, reduce the amount of I/O performed, and perform other important tasks. Scheduling where and when to run analytics is challenging, however. This paper quantifies the costs and benefits of different approaches to scheduling applications and analytics on nodes in large-scale applications, including space sharing, uncoordinated time sharing, and gang scheduled time sharing.
In next-generation extreme-scale systems, application performance will be limited by memory performance characteristics. The first exascale system is projected to contain many petabytes of memory. In addition to the sheer volume of the memory required, device trends, such as shrinking feature sizes and reduced supply voltages, have the potential to increase the frequency of memory errors. As a result, resilience to memory errors is a key challenge. In this paper, we evaluate the viability of using memory compression to repair detectable uncorrectable errors (DUEs) in memory. We develop a software library, evaluate its performance and demonstrate that it is able to significantly compress memory of HPC applications. Further, we show that exploiting compressed memory pages to correct memory errors can significantly improve application performance on next-generation systems.
Next-generation systems face a wide range of new potential sources of application interference, including resilience actions, system software adaptation, and in situ analytics programs. In this paper, we present a new model for analyzing the performance of bulk-synchronous HPC applications based on the use of extreme value theory. After validating this model against both synthetic and real applications, the paper then uses both simulation and modeling techniques to profile next-generation interference sources and characterize their behavior and performance impact on a selection of HPC benchmarks, mini-applications, and applications. Lastly, this work shows how the model can be used to understand how current interference mitigation techniques in multi-processors work.
Proceedings of the International Conference on Supercomputing
Fiala, David; Mueller, Frank; Ferreira, Kurt B.; Engelmann, Christian
Concern is growing in the high-performance computing (HPC) community on the reliability of future extreme- scale systems. Current efforts have focused on appli- cation fault-tolerance rather than the operating system (OS), despite the fact that recent studies have suggested that failures in OS memory may be more likely. The OS is critical to a system's correct and efficient operation of the node and processes it governs-and the parallel na- ture of HPC applications means any single node failure generally forces all processes of this application to ter- minate due to tight communication in HPC. Therefore, the OS itself must be capable of tolerating failures in a robust system. In this work, we introduce mini-ckpts, a framework which enables application survival despite the occurrence of a fatal OS failure or crash. minickpts achieves this tolerance by ensuring that the crit- ical data describing a process is preserved in persistent memory prior to the failure. Following the failure, the OS is rejuvenated via a warm reboot and the applica- tion continues execution effectively making the failure and restart transparent. The mini-ckpts rejuvenation and recovery process is measured to take between three to six seconds and has a failure-free overhead of between 3-5% for a number of key HPC workloads. In contrast to current fault-tolerance methods, this work ensures that the operating and runtime systems can continue in the presence of faults. This is a much finer-grained and dynamic method of fault-tolerance than the current coarse-grained application-centric methods. Handling faults at this level has the potential to greatly reduce overheads and enables mitigation of additional faults.
Fault tolerance is a key challenge to building the first exascale system. To understand the potential impacts of failures on next-generation systems, significant effort has been devoted to collecting, characterizing and analyzing failures on current systems. These studies require large volumes of data and complex analysis. Because the occurrence of failures in large-scale systems is unpredictable, failures are commonly modeled as a stochastic process. Failure data from current systems is examined in an attempt to identify the underlying probability distribution and its statistical properties. In this paper, we use modeling to examine the impact of failure distributions on the time-to-solution and the optimal checkpoint interval of applications that use coordinated checkpoint/restart. Using this approach, we show that as failures become more frequent, the failure distribution has a larger influence on application performance. We also show that as failure times are less tightly grouped (i.e., as the standard deviation increases) the underlying probability distribution has a greater impact on application performance. Finally, we show that computing the checkpoint interval based on the assumption that failures are exponentially distributed has a modest impact on application performance even when failures are drawn from a different distribution. Our work provides critical analysis and guidance to the process of analyzing failure data in the context of coordinated checkpoint/restart. Specifically, the data presented in this paper helps to distinguish cases where the failure distribution has a strong influence on application performance from those cases when the failure distribution has relatively little impact.
Baseman, Elisabeth B.; DeBardeleben, Nathan D.; Ferreira, Kurt B.; Levy, Scott L.; Raasch, Steven R.; Sridharan, Vilas S.; Siddiqua, Taniya S.; Guan, Qiang G.
Relaxed synchronization offers the potential for maintaining application scalability, by allowing many processes to make independent progress when some processes suffer delays. Yet the benefits of this approach for important parallel workloads have not been investigated in detail. In this paper, we use a validated simulation approach to explore the noise-mitigation effects of idealized nonblocking collectives, in workloads where these collectives are a major contributor to total execution time. Although nonblocking collectives are unlikely to provide significant noise mitigation to applications in the low operating system noise environments expected in next-generation high-performance computing systems, we show that they can potentially improve application runtime with respect to other noise types.