Technical Report Details
| Date |
27-5-2003 |
| Number |
DISI-TR-03-06 |
| Title |
Providing Single I/O Space and Multiple Fault Tolerance in a Distributed RAID |
| Authors |
Alessandro Di Marco, Giovanni Chiola, Giuseppe Ciaccio |
| Bibtex Entry |
@techreport{DRAIDreport,
author = {A. Di Marco AND G. Chiola AND G. Ciaccio},
title = {{Pro |
| E-mail |
ciaccio@disi.unige.it |
| Link |
ftp://ftp.disi.unige.it/person/CiaccioG/draid_techrep.ps.gz |
| Abstract |
Commodity EIDE disks provide low cost storage but are severely limited in
bandwidth and cannot be made fault-tolerant. On the other hand, conventional
RAID devices provide reliability and performance but worse price/performance
figures.
A cluster of PCs can be seen as a collection of networked low cost disks;
such a collection can be operated by proper software so as to provide the
abstraction of a single, larger block device. By adding suitable data
redundancy, such a disk collection as a whole could act as single, highly
fault tolerant, virtual RAID device, providing capacity and reliability
along with the convenient price/performance typical of commodity clusters.
Such a distributed RAID could even be accessible by each of the stations in the
cluster, rather than throughout one or few end-points, with a potentially
higher aggregate I/O bandwidh and better suitability to parallel I/O.
The overall picture is a prototype of a so-called ``I/O-centric cluster'',
namely, a cluster of PCs offering storage services under a Single I/O Space
(SIOS), rather than offering traditional computing power.
This paper reports about the design and performance of a prototype of such
a virtual RAID system, called DRAID, on a Gigabit Ethernet cluster of PCs. |
|
|
 |
