Open AccessProceedings Article
GPFS: A Shared-Disk File System for Large Computing Clusters
Frank B. Schmuck,Roger L. Haskin +1 more
- pp 231-244
Reads0
Chats0
TLDR
GPFS is IBM's parallel, shared-disk file system for cluster computers, available on the RS/6000 SP parallel supercomputer and on Linux clusters, and discusses how distributed locking and recovery techniques were extended to scale to large clusters.Citations
More filters
Proceedings ArticleDOI
A Technique for Lock-Less Mirroring in Parallel File Systems
TL;DR: This paper describes a data replication technique that meets the resiliency and consistency requirements of parallel file systems and provides scalable performance and measures the performance of the proposed mechanism by implementing it in a popular system.
Proceedings ArticleDOI
Performance impact of operating systems' caching parameters on parallel file systems
Eduardo C. Inacio,Mario A. R. Dantas,Francieli Zanon Boito,Philippe O. A. Navaux,Douglas Dyllon Jeronimo de Macedo +4 more
TL;DR: An analytical performance model is proposed in order to reflect caching effects on performance of file write operations and results indicate that the proposed model is effective in representing caching effects.
Proceedings ArticleDOI
vPFS+: Managing I/O Performance for Diverse HPC Applications
Ming Zhao,Yiqi Xu +1 more
TL;DR: In this article, a new proportional-share I/O scheduler, SFQ(D)+, is created to allow applications with various I/Os sizes and issue rates to share the storage with good application-level fairness and system-level utilization.
Proceedings ArticleDOI
Dynamic Provisioning of Storage Resources: A Case Study with Burst Buffers
TL;DR: This work proposes a proof-of-concept that is able to deploy, on-demand, a parallel file-system across intermediate storage nodes on a Cray XC50 system and shows how this mechanism can be easily extended to support more data managers and any type of intermediate storage.
Journal ArticleDOI
EML: An I/O scheduling algorithm in large-scale-application environments
TL;DR: An I/O scheduling algorithm, called EML (equal-length multi-level algorithm) is proposed to solve the problem of large amounts of data-intensive applications often access the data server concurrently, it can reduce the average response time significantly.
References
More filters
Book ChapterDOI
Notes on Data Base Operating Systems
TL;DR: This paper is a compendium of data base management operating systems folklore and focuses on particular issues unique to the transaction management component especially locking and recovery.
Proceedings ArticleDOI
Petal: distributed virtual disks
TL;DR: The design, implementation, and performance of Petal is described, a system that attempts to approximate this ideal in practice through a novel combination of features.
Journal ArticleDOI
Extendible hashing—a fast access method for dynamic files
TL;DR: This work studies, by analysis and simulation, the performance of extendible hashing and indicates that it provides an attractive alternative to other access methods, such as balanced trees.
Proceedings ArticleDOI
Frangipani: a scalable distributed file system
TL;DR: Initial measurements indicate that Frangipani has excellent single-server performance and scales well as servers are added, and can be exported to untrusted machines using ordinary network file access protocols.
Proceedings Article
Scalability in the XFS file system
TL;DR: The architecture and design of a new file system, XFS, for Silicon Graphics' IRIX operating system is described, and the use of B+ trees in place of many of the more traditional linear file system structures are discussed.