Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Performance Trace”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records

Chimbuko: A Workflow-Level Scalable Performance Trace Analysis Tool

ABSTRACT Due to the sheer volume of data it is typically impractical to analyze the detailed performance of an HPC application running at-scale. While conventional small-scale benchmarking and scaling studies are often sufficient for simple applications, many modern workflow-based applications couple multiple elements with competing resource demands and complex inter-communication patterns for which performance cannot easily be studied in isolation and at small scale. This work discusses Chimbuko, a performance analysis framework that provides real-time, in situ anomaly detection. By focusing specifically on performance anomalies and their origin (aka provenance), data volumes are dramatically reduced without losing necessary details. To the best of our knowledge, Chimbuko is the first online, distributed, and scalable workflow-level performance trace analysis framework. We demonstrate the tool's usefulness on Oak Ridge National Laboratory's Summit system.

97 MATHEMATICS AND COMPUTING↗

Chimbuko: A Workflow-Level Scalable Performance Trace Analysis Tool

Due to the sheer volume of data it is typically impractical to analyze the detailed performance of an HPC application running at-scale. While conventional small-scale benchmarking and scaling studies are often sufficient for simple applications, many modern workflow-based applications couple multiple elements with competing resource demands and complex inter-communication patterns for which performance cannot easily be studied in isolation and at small scale. This work discusses Chimbuko, a performance analysis framework that provides real-time, in situ anomaly detection. By focusing specifically on performance anomalies and their origin (aka provenance), data volumes are dramatically reduced without losing necessary details. To the best of our knowledge, Chimbuko is the first online, distributed, and scalable workflow-level performance trace analysis framework. We demonstrate the tool's usefulness on Oak Ridge National Laboratory's Summit system.

97 MATHEMATICS AND COMPUTING↗

Traveler: Navigating Task Parallel Traces for Performance Analysis

Understanding the behavior of software in execution is a key step in identifying and fixing performance issues. This is especially important in high performance computing contexts where even minor performance tweaks can translate into large savings in terms of computational resource use. To aid performance analysis, developers may collect an execution trace —a chronological log of program activity during execution. As traces represent the full history, developers can discover a wide array of possibly previously unknown performance issues, making them an important artifact for exploratory performance analysis. However, interactive trace visualization is difficult due to issues of data size and complexity of meaning. Traces represent nanosecond-level events across many parallel processes, meaning the collected data is often large and difficult to explore. The rise of asynchronous task parallel programming paradigms complicates the relation between events and their probable cause. Here, to address these challenges, we conduct a continuing design study in collaboration with high performance computing researchers. We develop diverse and hierarchical ways to navigate and represent execution trace data in support of their trace analysis tasks. Through an iterative design process, we developed Traveler , an integrated visualization platform for task parallel traces. Traveler provides multiple linked interfaces to help navigate trace data from multiple contexts. We evaluate the utility of Traveler through feedback from users and a case study, finding that integrating multiple modes of navigation in our design supported performance analysis tasks and led to the discovery of previously unknown behavior in a distributed array library.

97 MATHEMATICS AND COMPUTING↗

Signal Processing Based Method for Real-Time Anomaly Detection in High-Performance Computing

Performance anomalies can manifest as irregular execution times or abnormal execution events for many reasons, including network congestion and resource contention. Detecting such anomalies in real-time by analyzing the details of performance traces at scale is impractical due to the sheer volume of data High-Performance Computing (HPC) applications produce. In this paper, we propose formulating HPC performance anomaly detection as a signal-processing problem where anomalies can be treated as noise. We evaluate our proposed method in comparison with two other commonly used anomaly detection techniques of varying complexity based on their detection accuracy and scalability. Since real-time in-situ anomaly detection at a large scale requires lightweight methods that can handle a large volume of streaming data, we find that our proposed method provides the best trade-off. We then implement the proposed method in Chimbuko, the first online, distributed, and scalable workflow-level performance trace analysis framework. We compare our proposed signal-based anomaly detection algorithm with two other methods using a function of their accuracy, F1 score, and detection overhead. Our experiments demonstrate that our proposed approach achieves a 99% improvement for the benchmark datasets and a 93% improvement with Chimbuko traces.

99 GENERAL AND MISCELLANEOUS↗

Trace-based performance analysis for the petascale simulation code FLASH

Performance analysis of applications on modern high-end petascale systems is increasingly challenging due to the rising complexity and quantity of the computing units. This paper presents a performance-analysis study using the Vampir performance-analysis tool suite, which examines application behavior as well as the fundamental system properties. This study was carried out on the Jaguar system at Oak Ridge National Laboratory, the fastest computer on the November 2009 Top500 list. We analyzed the FLASH simulation code that is designed to be scaled with tens of thousands of CPU cores, which means that using existing performance-analysis tools is very complex. The study reveals two classes of performance problems that are relevant for very high CPU counts: MPI communication and scalable I/O. For both, solutions are presented and verified. Finally, the paper proposes improvements and extensions for event tracing tools in order to allow scalability of the tools towards higher degrees of parallelism.

Computer Science↗

Characterization of inexpensive metal oxide sensor performance for trace methane detection

Abstract. Methane, a major contributor to climate change, is emitted by a variety of natural and anthropogenic sources. Commercially available lab-grade instruments for sensing trace methane are expensive, and previous efforts to develop inexpensive, field-deployable trace methane sensors have had mixed results. Industrial and commercial metal oxide (MOx) methane sensors, which are intended for leak detection and safety monitoring, can potentially be repurposed and adapted for low-concentration sensing. As an initial step towards developing a low-cost sensing system, we characterize the performance of five off-the-shelf MOx sensors for 2–10 ppm methane detection in a laboratory setting (Figaro Engineering TGS2600, TGS2602, TGS2611-C00, TGS2611-E00, and Henan Hanwei Electronics MQ4). We identify TGS2611-C00, TGS2611-E00, and MQ4 as promising for trace methane sensing but show that variations in ambient humidity and temperature pose a challenge for the sensors in this application.

54 ENVIRONMENTAL SCIENCES↗

Reinforcement Learning for Load-balanced Parallel Particle Tracing

We explore an online reinforcement learning (RL) paradigm to dynamically optimize parallel particle tracing performance in distributed-memory systems. Our method combines three novel components: (1) a work donation algorithm, (2) a high-order workload estimation model, and (3) a communication cost model. First, we design an RL-based work donation algorithm. Our algorithm monitors workloads of processes and creates RL agents to donate data blocks and particles from high-workload processes to low-workload processes to minimize program execution time. The agents learn the donation strategy on the fly based on reward and cost functions designed to consider processes' workload changes and data transfer costs of donation actions. Second, we propose a workload estimation model, helping RL agents estimate the workload distribution of processes in future computations. Third, we design a communication cost model that considers both block and particle data exchange costs, helping RL agents make effective decisions with minimized communication costs. We demonstrate that our algorithm adapts to different flow behaviors in large-scale fluid dynamics, ocean, and weather simulation data. Our algorithm improves parallel particle tracing performance in terms of parallel efficiency, load balance, and costs of I/O and communication for evaluations with up to 16,384 processors.

Distributed and parallel particle tracing↗

Method of resource-limited device and device class identification using system and function call tracing techniques, performance, and statistical analysis

Methods for cyber physical systems device classification are provided. A method can include receiving system and function calls and parameters and a device performance index from an unknown CPS device and a device performance index of similar class of CPS devices, calculating an autocorrelation value between different realizations of the system and function calls and parameters of the known CPS device, determining whether the autocorrelation value is greater than a threshold amount, and storing the system and function calls and parameters and the device performance characteristics of the known CPS device in the database. A method can also include calculating a correlation between system and function calls and parameters of an unknown CPS device and known CPS devices classes included in the database, as well as determining whether the maximum correlation is also greater than a threshold amount.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Computational Performance Bounds Prediction in Quantum Computing With Unstable Noise

Quantum computing has significantly advanced in recent years, boasting devices with hundreds of quantum bits (qubits), hinting at its potential quantum advantage over classical computing. Yet, noise in quantum devices poses significant barriers to realizing this supremacy. Understanding noise’s impact is crucial for reproducibility and application reuse; moreover, the next-generation quantum-centric supercomputing essentially requires efficient and accurate noise characterization to support system management (e.g., job scheduling), where ensuring correct functional performance (i.e., fidelity) of jobs on available quantum devices can even be higher-priority than traditional objectives. However, noise fluctuates over time, even on the same quantum device, which makes predicting the computational bounds for on-the-fly noise is vital. Noisy quantum simulation can offer insights but faces efficiency and scalability issues. Here, in this work, we propose a data-driven workflow, namely QuBound, to predict computational performance bounds. It decomposes historical performance traces to isolate noise sources and devises a novel encoder to embed circuit and noise information processed by a Long Short-Term Memory (LSTM) network. For evaluation, we compare QuBound with a state-of-the-art learning-based predictor, which only generates a single performance value instead of a bound. Experimental results show that the result of the existing approach falls outside of performance bounds, while all predictions from our QuBound with the assistance of performance decomposition better fit the bounds. Moreover, QuBound can efficiently produce practical bounds for various circuits with over 106 speedup over simulation; in addition, the range from QuBound is over 10× narrower than the state-of-the-art analytical approach.

Li, Jinyang [George Mason Univ., Fairfax, VA (Unit↗

Panorama 360 (Final Report)

This is the final technical report for the DOE-funded Panorama 360 project. Panorama 360 provided a resource for the collection, analysis, and sharing of performance data about end-to-end scientific workflows executing on DOE facilities. The work focused on workflows that include experimental data generation at DOE facilities. The main activities of Panorama 360 include the development of: 1. A distributed repository that stores different types of workflow execution data (e.g., point and time series performance traces at fine- and coarse-grained levels); 2. A set of open-source data capture, curation, and publishing tools fully integrated with a state-of-the-art workflow management system that automates data ingestion to the repository and enables users to discover, query, and process data from the repository; 3. A set of analysis algorithms and machine learning based tools to perform analysis and characterization of the gathered data, which can be used to detect anomalous performance or system faults; and 4. Best practices and recommendations for workflow evaluation, analysis, execution, and architectures.

97 MATHEMATICS AND COMPUTING↗

Panorama 360 (Final Report)

This final technical report from the lead institution, USC grant #DE-SC0012636, serves as the final technical report for collaborative institution UNC-CH grant #DE-SC0012390. The goal was to develop a repository and associated capabilities for data collection, ingestion, and analysis for a broad class of DOE applications that span experimental and simulation science workflows. In particular, this work focuses on workflows that include experimental data generation at DOE facilities. The main activities of Panorama 360 include the development of: (1) A distributed repository that stores different types of workflow execution data (e.g., point and time series performance traces at fine- and coarse-grained levels); (2) A set of open-source data capture, curation, and publishing tools fully integrated with a state-of-the-art workflow management system that automates data ingestion to the repository and enables users to discover, query, and process data from the repository; (3) A set of analysis algorithms and machine learning based tools to perform analysis and characterization of the gathered data, which can be used to detect anomalous performance or system faults; and (4) Best practices and recommendations for workflow evaluation, analysis, execution, and architectures.

97 MATHEMATICS AND COMPUTING↗

Composition, Emissions, and Air Quality Impacts of Hazardous Air Pollutants in Unburned Natural Gas from Residential Stoves in California

The presence of hazardous air pollutants (HAPs) entrained in end-use natural gas (NG) is an understudied source of human health risks. We performed trace gas analyses on 185 unburned NG samples collected from 159 unique residential NG stoves across seven geographic regions in California. Our analyses commonly detected 12 HAPs with significant variability across region and gas utility. Mean regional benzene, toluene, ethylbenzene, and total xylenes (BTEX) concentrations in end-use NG ranged from 1.6–25 ppmv–benzene alone was detected in 99% of samples, and mean concentrations ranged from 0.7–12 ppmv (max: 66 ppmv). By applying previously reported NG and methane emission rates throughout California’s transmission, storage, and distribution systems, we estimated statewide benzene emissions of 4,200 (95% CI: 1,800–9,700) kg yr –1 that are currently not included in any statewide inventories–equal to the annual benzene emissions from nearly 60,000 light-duty gasoline vehicles. Additionally, we found that NG leakage from stoves and ovens while not in use can result in indoor benzene concentrations that can exceed the California Office of Environmental Health Hazard Assessment 8-h Reference Exposure Level of 0.94 ppbv–benzene concentrations comparable to environmental tobacco smoke. This study supports the need to further improve our understanding of leaked downstream NG as a source of health risk.

42 ENGINEERING↗

An optimal merging technique for high-resolution precipitation products: OPTIMAL MERGING OF PRECIPITATION METHOD

Precipitation products are currently available from various sources at higher spatial and temporal resolution than any time in the past. Each of the precipitation products has its strengths and weaknesses in availability, accuracy, resolution, retrieval techniques and quality control. By merging the precipitation data obtained from multiple sources, one can improve its information content by minimizing these issues. However, precipitation data merging poses challenges of scale-mismatch, and accurate error and bias assessment. In this paper we present Optimal Merging of Precipitation (OMP), a new method to merge precipitation data from multiple sources that are of different spatial and temporal resolutions and accuracies. This method is a combination of scale conversion and merging weight optimization, involving performance-tracing based on Bayesian statistics and trend-analysis, which yields merging weights for each precipitation data source. The weights are optimized at multiple scales to facilitate multiscale merging and better precipitation downscaling. Precipitation data used in the experiment include products from the 12-km resolution North American Land Data Assimilation (NLDAS) system, the 8-km resolution CMORPH and the 4-km resolution National Stage-IV QPE. The test cases demonstrate that the OMP method is capable of identifying a better data source and allocating a higher priority for them in the merging procedure, dynamically over the region and time period. This method is also effective in filtering out poor quality data introduced into the merging process.

54 ENVIRONMENTAL SCIENCES↗

PLAStiCC: Predictive Look-Ahead Scheduling for Continuous dataflows on Clouds

Scalable stream processing and continuous dataflow systems are gaining traction with the rise of big data due to the need for processing high velocity data in near real time. Unlike batch processing systems such as MapReduce and workflows, static scheduling strategies fall short for continuous dataflows due to the variations in the input data rates and the need for sustained throughput. The elastic resource provisioning of cloud infrastructure is valuable to meet the changing resource needs of such continuous applications. However, multi-tenant cloud resources introduce yet another dimension of performance variability that impacts the application’s throughput. In this paper we propose PLAStiCC, an adaptive scheduling algorithm that balances resource cost and application throughput using a prediction-based look-ahead approach. It not only addresses variations in the input data rates but also the underlying cloud infrastructure. In addition, we also propose several simpler static scheduling heuristics that operate in the absence of accurate performance prediction model. These static and adaptive heuristics are evaluated through extensive simulations using performance traces obtained from public and private IaaS clouds. Our results show an improvement of up to 20% in the overall profit as compared to the reactive adaptation algorithm.

97 MATHEMATICS AND COMPUTING↗