Scalability and Failure Recovery in a Linux Cluster File System

Abstract: 

In this paper we describe how we implemented journaling and recovery in the Global File System (GFS), a shared-disk, cluster file system for Linux. We also present our latest performance results for a 16-way Linux cluster.

Traditional local file systems support a persistent name space by creating a mapping between blocks found on disk drives and a set of files, file names, and directories. These file systems view devices as local: devices are not shared so there is no need in the file system to enforce device sharing semantics. Instead, the focus is on aggressively caching and aggregating file system operations to improve performance by reducing the number of actual disk accesses required for each file system operation.

BibTeX
@inproceedings {271221,
author = {Kenneth Preslan and Andrew Barry and Jonathan Brassow and Michael Declerck and A.J. Lewis and Adam Manthei and Ben Marzinski and Erling Nygaard and Seth Van Oort and David Teigland and Mike Tilstra and Steve Whitehouse and Matthew O{\textquoteright}Keefe},
title = {Scalability and Failure Recovery in a Linux Cluster File System},
booktitle = {4th Annual Linux Showcase \& Conference (ALS 2000)},
year = {2000},
address = {Atlanta, GA },
url = {https://www.usenix.org/conference/als-2000/scalability-and-failure-recovery-linux-cluster-file-system},
publisher = {USENIX Association},
month = oct
}
Download

Links

Paper: 
http://www.usenix.org/publications/library/proceedings/als00/2000papers/papers/full_papers/preslan/preslan.pdf
Paper (HTML): 
http://www.usenix.org/publications/library/proceedings/als00/2000papers/papers/full_papers/preslan/preslan_html/index.html