Scientific workflows drive most modern large-scale science breakthroughs by allowing scientists to define their computations as a set of jobs executed in a given order based on their data dependencies. Workflow management systems (WMSs) have become key to automating scientific workflows-executing computational jobs and orchestrating data transfers between those jobs running on complex high-performance computing (HPC) platforms. Traditionally, WMSs use files to communicate between jobs: a job writes out files that are read by other jobs. However, HPC machines face a growing gap between their storage and compute capabilities. To address that concern, the scientific community has adopted a new approach called in situ, which bypasses costly parallel filesystem I/O operations with faster in-memory or in-network communications. When using in situ approaches, communication and computations can be interleaved. In this work, we leverage the Decaf in situ dataflow framework to accelerate task-based scientific workflows managed by the Pegasus WMS, by replacing file communications with faster MPI messaging. We propose a new execution engine that uses Decaf to manage communications within a sub-workflow (i.e., set of jobs) to optimize inter-job communications. We consider two workflows in this study: (i) a synthetic workflow that benchmarks and compares file- and MPI-based communication; and (ii) a realistic bioinformatics workflow that computes mu-tational overlaps in the human genome. Experiments show that in situ communication can improve the bioinformatics workflow execution time by 22% to 30% compared with file communication. Our results motivate further opportunities and challenges for bridging traditional WMSs with in situ frameworks.
more »
« less
This content will become publicly available on May 1, 2024
Evaluating Asynchronous Parallel I/O on HPC Systems
Parallel I/O is an effective method to optimize data movement between memory and storage for many scientific applications. Poor performance of traditional disk-based file systems has led to the design of I/O libraries which take advantage of faster memory layers, such as on-node memory, present in high-performance computing (HPC) systems. By allowing caching and prefetching of data for applications alternating computation and I/O phases, a faster memory layer also provides opportunities for hiding the latency of I/O phases by overlapping them with computation phases, a technique called asynchronous I/O. Since asynchronous parallel I/O in HPC systems is still in the initial stages of development, there hasn't been a systematic study of the factors affecting its performance.In this paper, we perform a systematic study of various factors affecting the performance and efficacy of asynchronous I/O, we develop a performance model to estimate the aggregate I/O bandwidth achievable by iterative applications using synchronous and asynchronous I/O based on past observations, and we evaluate the performance of the recently developed asynchronous I/O feature of a parallel I/O library (HDF5) using benchmarks and real-world science applications. Our study covers parallel file systems on two large-scale HPC systems: Summit and Cori, the former with a GPFS storage and the latter with a Lustre parallel file system.
more »
« less
- Award ID(s):
- 1812727
- NSF-PAR ID:
- 10437083
- Date Published:
- Journal Name:
- 10.1109/IPDPS54959.2023.00030
- Page Range / eLocation ID:
- 211 to 221
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Parallel File Systems (PFSs) are frequently deployed on leadership High Performance Computing (HPC) systems to ensure efficient I/O, persistent storage and scalable performance. Emerging Deep Learning (DL) applications incur new I/O and storage requirements to HPC systems with batched input of small random files. This mandates PFSs to have commensurate features that can meet the needs of DL applications. BeeGFS is a recently emerging PFS that has grabbed the attention of the research and industry world because of its performance, scalability and ease of use. While emphasizing a systematic performance analysis of BeeGFS, in this paper, we present the architectural and system features of BeeGFS, and perform an experimental evaluation using cutting-edge I/O, Metadata and DL application benchmarks. Particularly, we have utilized AlexNet and ResNet-50 models for the classification of ImageNet dataset using the Livermore Big Artificial Neural Network Toolkit (LBANN), and ImageNet data reader pipeline atop TensorFlow and Horovod. Through extensive performance characterization of BeeGFS, our study provides a useful documentation on how to leverage BeeGFS for the emerging DL applications.more » « less
-
Many applications are increasingly becoming I/O-bound. To improve scalability, analytical models of parallel I/O performance are often consulted to determine possible I/O optimizations. However, I/O performance modeling has predominantly focused on applications that directly issue I/O requests to a parallel file system or a local storage device. These I/O models are not directly usable by applications that access data through standardized I/O libraries, such as HDF5, FITS, and NetCDF, because a single I/O request to an object can trigger a cascade of I/O operations to different storage blocks. The I/O performance characteristics of applications that rely on these libraries is a complex function of the underlying data storage model, user-configurable parameters and object-level access patterns. As a consequence, I/O optimization is predominantly an ad-hoc process that is performed by application developers, who are often domain scientists with limited desire to delve into nuances of the storage hierarchy of modern computers.This paper presents an analytical cost model to predict the end-to-end execution time of applications that perform I/O through established array management libraries. The paper focuses on the HDF5 and Zarr array libraries, as examples of I/O libraries with radically different storage models: HDF5 stores every object in one file, while Zarr creates multiple files to store different objects. We find that accessing array objects via these I/O libraries introduces new overheads and optimizations. Specifically, in addition to I/O time, it is crucial to model the cost of transforming data to a particular storage layout (memory copy cost), as well as model the benefit of accessing a software cache. We evaluate the model on real applications that process observations (neuroscience) and simulation results (plasma physics). The evaluation on three HPC clusters reveals that I/O accounts for as little as 10% of the execution time in some cases, and hence models that only focus on I/O performance cannot accurately capture the performance of applications that use standard array storage libraries. In parallel experiments, our model correctly predicts the fastest storage library between HDF5 and Zarr 94% of the time, in contrast with 70% of the time for a cutting-edge I/O model.more » « less
-
On large-scale high performance computing (HPC) systems, applications are provisioned with aggregated resources to meet their peak demands for brief periods. This results in resource underutilization because application requirements vary a lot during execution. This problem is particularly pronounced for deep learning applications that are running on leadership HPC systems with a large pool of burst buffers in the form of flash or non-volatile memory (NVM) devices. In this paper, we examine the I/O patterns of deep neural networks and reveal their critical need of loading many small samples randomly for successful training. We have designed a specialized Deep Learning File System (DLFS) that provides a thin set of APIs. Particularly, we design the metadata management of DLFS through an in-memory tree-based sample directory and its file services through the user-level SPDK protocol that can disaggregate the capabilities of NVM Express (NVMe) devices to parallel training tasks. Our experimental results show that DLFS can dramatically improve the throughput of training for deep neural networks on NVMe over Fabric, compared with the kernel-based Ext4 file system. Furthermore, DLFS achieves efficient user-level storage disaggregation with very little CPU utilization.more » « less
-
Large-scale parallel file systems (PFSs) play an essential role in high-performance computing (HPC). However, despite their importance, their reliability is much less studied or understood compared with that of local storage systems or cloud storage systems. Recent failure incidents at real HPC centers have exposed the latent defects in PFS clusters as well as the urgent need for a systematic analysis. To address the challenge, we perform a study of the failure recovery and logging mechanisms of PFSs in this article. First, to trigger the failure recovery and logging operations of the target PFS, we introduce a black-box fault injection tool called PFault , which is transparent to PFSs and easy to deploy in practice. PFault emulates the failure state of individual storage nodes in the PFS based on a set of pre-defined fault models and enables examining the PFS behavior under fault systematically. Next, we apply PFault to study two widely used PFSs: Lustre and BeeGFS. Our analysis reveals the unique failure recovery and logging patterns of the target PFSs and identifies multiple cases where the PFSs are imperfect in terms of failure handling. For example, Lustre includes a recovery component called LFSCK to detect and fix PFS-level inconsistencies, but we find that LFSCK itself may hang or trigger kernel panics when scanning a corrupted Lustre. Even after the recovery attempt of LFSCK, the subsequent workloads applied to Lustre may still behave abnormally (e.g., hang or report I/O errors). Similar issues have also been observed in BeeGFS and its recovery component BeeGFS-FSCK. We analyze the root causes of the abnormal symptoms observed in depth, which has led to a new patch set to be merged into the coming Lustre release. In addition, we characterize the extensive logs generated in the experiments in detail and identify the unique patterns and limitations of PFSs in terms of failure logging. We hope this study and the resulting tool and dataset can facilitate follow-up research in the communities and help improve PFSs for reliable high-performance computing.more » « less
