A Decentralized Algorithm for Erasure-Coded Virtual Disks

Authors:
Svend Frolund;Arif Merchant;Yasushi Saito;Susan Spence;Alistair Veitch
Affiliations:
HP Labs, Palo Alto, CA;HP Labs, Palo Alto, CA;HP Labs, Palo Alto, CA;HP Labs, Palo Alto, CA;HP Labs, Palo Alto, CA
Venue:
DSN '04 Proceedings of the 2004 International Conference on Dependable Systems and Networks
Year:
2004

Citing 0
Cited 9

FAB: building distributed enterprise disk arrays from commodity components

ASPLOS XI Proceedings of the 11th international conference on Architectural support for programming languages and operating systems
Efficient Updates in Highly Available Distributed Random Access Memory

ICPADS '06 Proceedings of the 12th International Conference on Parallel and Distributed Systems - Volume 1
Optimistic Erasure-Coded Distributed Storage

DISC '08 Proceedings of the 22nd international symposium on Distributed Computing
GRID codes: Strip-based erasure codes with high fault tolerance for storage systems

ACM Transactions on Storage (TOS)
Remote storage with byzantine servers

Proceedings of the twenty-first annual symposium on Parallelism in algorithms and architectures
Maintaining and checking parity in highly available Scalable Distributed Data Structures

Journal of Systems and Software
Consistency and fault tolerance for erasure-coded distributed storage systems

Proceedings of the fifth international workshop on Data-Intensive Distributed Computing Date
Analysis for REPERA: A Hybrid Data Protection Mechanism in Distributed Environment

International Journal of Cloud Applications and Computing
Parity logging with reserved space: towards efficient updates and recovery in erasure-coded clustered storage

FAST'14 Proceedings of the 12th USENIX conference on File and Storage Technologies

Quantified Score

Hi-index	0.00

Visualization

Abstract

A Federated Array of Bricks is a scalable distributedstorage system composed from inexpensive storage bricks.It achieves high reliability with low cost by using erasurecoding across the bricks to maintain data reliability inthe face of brick failures. Erasure coding generates n encodedblocks from m data blocks (n m) and permits thedata blocks to be reconstructed from any m of these encodedblocks. We present a new fully decentralized erasure-codingalgorithm for an asynchronous distributed system.Our algorithm provides fully linearizable read-write accessto erasure-coded data and supports concurrent I/O controllersthat may crash and recover. Our algorithm relieson a novel quorum construction where any two quorums intersectin m processes.