ParaLog: Consistent Host-side Logging for Parallel Checkpoints

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chien, Steven W. D., Sato, Kento, Podobas, Artur, Jansson, Niclas, Markidis, Stefano, Honda, Michio
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909861665570816
author Chien, Steven W. D.
Sato, Kento
Podobas, Artur
Jansson, Niclas
Markidis, Stefano
Honda, Michio
author_facet Chien, Steven W. D.
Sato, Kento
Podobas, Artur
Jansson, Niclas
Markidis, Stefano
Honda, Michio
contents Output-intensive scientific applications are highly sensitive to low storage throughput. While existing scientific application stacks are optimized for traditional High-Performance Computing (HPC) environments with high remote storage and network bandwidth, these assumptions often fail in modern settings like cloud deployment. This is because the existing scientific application I/O stack fails to leverage the available resources. At the same time, scientific applications exhibit special synchronization and data output requirements that are difficult to satisfy using traditional approaches such as block-level or filesystem-level caching. We introduce ParaLog, a distributed host-side logging approach designed to accelerate scientific applications transparently. ParaLog emphasizes deployability, enabling support for unmodified message passing interface (MPI) applications and implementations while preserving crash consistency semantics. We evaluate ParaLog across traditional HPC, cloud HPC, local clusters, and hybrid environments, demonstrating its capability to reduce end-to-end execution time by 13-26% for popular scientific applications in cloud settings.
format Preprint
id arxiv_https___arxiv_org_abs_2401_14576
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle ParaLog: Consistent Host-side Logging for Parallel Checkpoints
Chien, Steven W. D.
Sato, Kento
Podobas, Artur
Jansson, Niclas
Markidis, Stefano
Honda, Michio
Distributed, Parallel, and Cluster Computing
Performance
Output-intensive scientific applications are highly sensitive to low storage throughput. While existing scientific application stacks are optimized for traditional High-Performance Computing (HPC) environments with high remote storage and network bandwidth, these assumptions often fail in modern settings like cloud deployment. This is because the existing scientific application I/O stack fails to leverage the available resources. At the same time, scientific applications exhibit special synchronization and data output requirements that are difficult to satisfy using traditional approaches such as block-level or filesystem-level caching. We introduce ParaLog, a distributed host-side logging approach designed to accelerate scientific applications transparently. ParaLog emphasizes deployability, enabling support for unmodified message passing interface (MPI) applications and implementations while preserving crash consistency semantics. We evaluate ParaLog across traditional HPC, cloud HPC, local clusters, and hybrid environments, demonstrating its capability to reduce end-to-end execution time by 13-26% for popular scientific applications in cloud settings.
title ParaLog: Consistent Host-side Logging for Parallel Checkpoints
topic Distributed, Parallel, and Cluster Computing
Performance
url https://arxiv.org/abs/2401.14576