Distributed RAID-a new multiple copy algorithm

Stonebraker, Michael; Schloss, G.A.

doi:10.1109/icde.1990.113496

Cited by 67 publications

(20 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At the lowest level, performance improvements are being achieved at the hardware and network level. Fine-grain parallelism at the disk level has been proposed through mechanisms s u c h as disk striping, interleaving, RAID and RADD 28,31]. Finally, to support solutions to the I/O problem, new disk architectures must be su ciently exible and programmable that new I/O paradigms can be implemented and tested.…”

Section: Introductionmentioning

confidence: 99%

Distributed scheduling algorithms to improve the performance of parallel data transfers

Durand¹,

Jain²,

Tseytlin³

1994

SIGARCH Comput. Archit. News

View full text Add to dashboard Cite

The cost of data transfers, and in particular of I/O operations, is a growing problem in parallel computing. This performance bottleneck is especially severe for data-intensive a p p l ications such a s m ultimedia information systems, databases, and Grand Challenge problems. A promising approach to alleviating this bottleneck i s t o s c hedule parallel I/O operations explicitly.Although centralized algorithms for batch s c heduling of parallel I/O operations have previously been developed, they are not be appropriate for all applications and architectures. We develop a class of decentralized algorithms for scheduling parallel I/O operations, where the objective is to reduce the time required to complete a given set of transfers. These algorithms, based on edge-coloring and matching of bipartite graphs, rely upon simple heuristics to obtain shorter schedules. We present s i m ulation results indicating that the best of our algorithms can produce schedules whose length is within 2 -20% of the optimal schedule, a substantial improvement on previous decentralized algorithms. We discuss theoretical and experimental work in progress and possible extensions.

show abstract

Section: Introductionmentioning

confidence: 99%

Distributed scheduling algorithms to improve the performance of parallel data transfers

Durand¹,

Jain²,

Tseytlin³

1994

SIGARCH Comput. Archit. News

View full text Add to dashboard Cite

show abstract

“…A unique advantage of this RAID-5 is that it does not suffer from the short-write problem [8,9] that plagues all software RAID implementations; from a storage network viewpoint of computing the parity, writing a single octet is almost as inexpensive as writing an entire disk block. Writes shorter than 512 bytes sectors are implemented as a read-modify-write operation by the FlowStack core in the disk block servers, but this is a fast operation as it operates out of on-disk buffers and over a dedicated IDE/SATA interface with spare capacity.…”

Section: Raid5 Operationmentioning

confidence: 99%

De-layered grid storage server

Shrikumar¹

2005

SIGBED Rev.

View full text Add to dashboard Cite

show abstract

“…The distributed RAID concept was proposed by Stonebraker and Schloss [1]. Examples of early-distributed RAID systems include Swift/RAID [2], Petal [3] and Tertiary Disk [4].…”

Section: Introductionmentioning

confidence: 99%

SARCNFS: Self-Adaptive Redundancy Clustered NAS File System

Cai

Xie

Ren

et al. 2004

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this paper, we describe the design and implementation of SARCNFS File System for network-attached clustered storage system. SARCNFS stripes the data and metadata among multiple NAS nodes, and provides file redundancy scheme and synchronization mechanism for distributed RAID5. SARCNFS uses a self-adaptive redundancy scheme for file data accesses that uses RAID5-level for large writes, and RAID1-level for small write so as to dynamically provide flexible switch between RAID1 and RAID5 to provide the best performance. In addition, SARCNFS proposed a simple distributed locking mechanism that uses RAID5-level for full stripe writes, and RAID1-level for temporary storing data from partial stripe updates. As a result, low response latency, high performance and strong reliability are achieved.

show abstract

Distributed RAID-a new multiple copy algorithm

Cited by 67 publications

References 11 publications

Distributed scheduling algorithms to improve the performance of parallel data transfers

Distributed scheduling algorithms to improve the performance of parallel data transfers

De-layered grid storage server

SARCNFS: Self-Adaptive Redundancy Clustered NAS File System

Contact Info

Product

Resources

About