RAID: high-performance, reliable secondary storage

@article{Chen1994RAIDHR,
  title={RAID: high-performance, reliable secondary storage},
  author={Peter M. Chen and Edward K. Lee and Garth A. Gibson and Randy H. Katz and David A. Patterson},
  journal={ACM Comput. Surv.},
  year={1994},
  volume={26},
  pages={145-185}
}
Disk arrays were proposed in the 1980s as a way to use parallelism between multiple disks to improve aggregate I/O performance. Today they appear in the product lines of most major computer manufacturers. This article gives a comprehensive overview of disk arrays and provides a framework in which to organize current and future work. First, the article introduces disk technology and reviews the driving forces that have popularized disk arrays: performance and reliability. It discusses the two… 

Multi-Partition RAID: A New Method for Improving Performance of Disk Arrays under Failure

TLDR
A new variation of RAID organization, multi-partition RAID (mP-RAID), is proposed to improve storage ef"ciency and reduce performance degradation when disk failures occur.

Redundant array of inexpensive disks (RAID)

TLDR
Slight modifications to these levels have recently resulted in the specification of levels 6 and 7 of RAID, which differ in the way they provide for speed and/or reliability.

High-reliability high-performance optical data storage system architecture

TLDR
This paper proposes a high performance and high reliability optical disk array architecture with less redundancy, called Mirror Striped Disk Array (MSDA), which can achieve the high reliability and high performance optical storage system without adding any extra redundancy and without losing any performance compared with RAID level 0 architecture.

Studies of disk arrays tolerating two disk failures and a proposal for a heterogeneous disk array

TLDR
The second part of this dissertation describes a new disk array architecture, namely Heterogeneous Disk Array (HDA), motivated by a few observations of the trends in storage technology.

Multi-level RAID for very large disk arrays

TLDR
The reliability analysis of Markov chain models of VLDAs of Very Large Disk Arrays is outlined and conclusions from simulation results are reported.

Mirrored and Hybrid Disk Arrays: Organization, Scheduling, Reliability, and Performance

TLDR
The reliability of mirrored disk organizations against each other and hybrid disks and erasure coded disk arrays are compared and hybrid disk arrays which store XORed instead of replicated data tend to have a higher reliability than mirrored disks, but incur a higher overhead in updating data.

Striping Policies in Multiclass Disk Arrays Ph.d. Thesis Proposal

TLDR
The goal of this proposed research is to determine striping policies for heterogeneous workloads providing both high I/O rates and high data rates through analytic modeling, simulations, and trace data executions.

Mirrored and hybrid disk arrays and their reliability

TLDR
The application of the asymptotic expansion method to hierarchical RAID shows that it is advantageous to associate higher redundancy with lower levels at the same overall redundancy overhead, and it is shown that sharing disk space sharing between RAID1 and RAID5 in heterogeneous disk arrays—HDAs may result in a lowered reliability.

A Study on Architectures of RAID Level 4 Disk Arrays

TLDR
It is concluded that the extended and dual parity methods forlevel 4 can stand comparison with sparing methods for level 5 in normal mode of small arrays, and also have the advantage of the recovery overhead.

Performance vs. cost of redundant arrays of inexpensive disks

...

References

SHOWING 1-10 OF 101 REFERENCES

Redundant disk arrays - reliable, parallel secondary storage

TLDR
This dissertation presents analytic models for disk-array lifetime, evaluates these against event-driven simulation, and applies them to an example redundant disk array, showing that a 10% overhead for an N + 1-parity encoding plus a 10%, overhead for on-line spares can provide higher reliability than the 100% overhead of conventional mirrored disks.

How reliable is a RAID?

TLDR
The authors present a scheme in which the system support components are organized into groups orthogonal to the data redundancy groups, thus guaranteeing that no single disk or component failure will permanently lose data.

Performance Analysis of Disk Arrays under Failure

TLDR
A new variation of the RAID organization is proposed that has significant advantages in both reducing the magnitude of the performance degradation when there is a single failure and can also reduce the mean time to system failure.

Performance of a disk array protype

TLDR
Examination of the performance limits of each component of the array usiug SCSI bus traces, Sprite operating system traces and user programs concludes that merely using a powerful host CPU and many disks will not provide the full bandwidth possible from disk arrays.

Performance consequences of parity placement in disk arrays

TLDR
It is shown that for relatively large request sizes of hundreds of kilobytes, the choice of parity placement significantly affects performance and the way parity is distributed in a RAID has significant consequences for performance.

Striping in a RAID level 5 disk array

TLDR
It is recommended (in the absence of specific workload information) that the striping unit over a wide range of RAID Level 5 disk array sizes be equal to 1/2 * average positioning time * disk transfer rate.

Maintaining good performance in disk arrays during failure via uniform parity group distribution

  • S. NgR. Mattson
  • Computer Science
    Proceedings of the First International Symposium on High-Performance Distributed Computing. (HPDC-1)
  • 1992
TLDR
This paper describes techniques that will enable the disk array to maintain good performance in the event of a disk failure, and shows how this sequential process can be broken down into multiple parallel processes distributed throughout the array, thus shortening the reconstruction time.

Fast, on-line failure recovery in redundant disk arrays

TLDR
An implementation of disk-oriented reconstruction, a data recovery algorithm that allows the reconstruction process to absorb essentially all the disk bandwidth not consumed by the user processes, is presented, and this algorithm is compared to a previous-proposed parallel stripe-oriented approach.

RAID-II: a high-bandwidth network file server

TLDR
RAID-II runs Log-Structured File System (LFS) software to optimize performance for bandwidth-intensive applications and a preliminary implementation of LFS on RAID-II delivers 21 megabyte/second on large read requests and 15 megabytes/ second on large write operations.

Crosshatch disk array for improved reliability and performance

  • S. Ng
  • Computer Science
    Proceedings of 21 International Symposium on Computer Architecture
  • 1994
TLDR
The author presents a new array architecture which outperforms other currently known array architectures when some supporting hardware such as a controller or cable has failed and is, therefore, a more reliable and robust architecture.
...