Electrical computers and digital processing systems: memory – Storage accessing and control – Specific memory composition
Reexamination Certificate
1999-05-19
2003-06-17
Bragdon, Reginald G. (Department: 2186)
Electrical computers and digital processing systems: memory
Storage accessing and control
Specific memory composition
C711S165000, C711S172000, C714S006130
Reexamination Certificate
active
06581135
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an information storage system and, more particularly, to an information storage system including a plurality of information storage devices each of which stores a part of information so that the entire information is stored as a whole.
2. Description of the Related Art
In recent years, a demand for a high-speed transfer of a large amount of sound or image data has been increased in relation to the rapid growth of internet communications via personal computers. However, peripheral systems such as a network for smoothly transferring and storing a large amount of data has not been sufficiently developed. Additionally, a network communication protocol technology for transmitting such a large amount of data in real time has not been sufficiently developed. Thus, it is desirous to urgently develop a system that can transfer a large amount of data at a high-speed by using existing techniques.
A system referred to as a redundant array of inexpensive disks (RAID) has been developed for retrieving and reproducing image data from an information storage apparatus having a high-speed transfer function, such as a non-linear editing of a video on demand (VOD) or a video image processing. The RAID system was revealed in the thesis titled “A Case for Redundant Arrays of Inexpensive Disks” written by David A. Peterson, Garth Gibson and Randy H. Katz, which thesis was published in 1987 the University of California. The RAID system uses an array of a plurality of hard disk drives so as to store a set of data by dividing the data into a plurality of data blocks. The data blocks are stored concurrently in the hard disk drives. Thereby, the RAID system can achieve a high-speed data transfer.
A description will now be given of the RAID system in detail.
FIG. 1
is an illustration for explaining a system structure of the RAID system.
The RAID system
1
comprises a host computer
2
and a plurality of external information-storage devices
3
-
1
to
3
-n.
When data is stored by the RAID system, there is a method in which the same contents are written in the plurality of external information-storage devices. Additionally, there is a method in which a set of data to be stored is divided into a plurality of data blocks (referred to as stripping), and data blocks are stored in the external information-storage devices on an individual data block basis. In this method, an error correction code is produced from the plurality of data blocks and stored in the external information-storage devices so as to improve data reliability. Accordingly, if one of the data blocks is lost, correct data can be restored according to the remaining data blocks and the error correction code.
It should be noted that there are six types, RAID-0 to RAID-5, the RAID system due to differences in dividing methods of a set of data and methods for managing the error correction code.
A description will now be given of RAID-0. In RAID-0, a set of data is divided or stripped into a plurality of data blocks. The data blocks are stored in different external information-storage devices, respectively. It should be noted that since the error correction code is not produced in RAID-0, there is no fault tolerance.
In RAID-1, two external information-storage devices are provided. A set of data is stored in one of the external information-storage devices, and the same contents are stored in the other. Accordingly, if one of the external information-storage devices fails and input and output operations cannot be performed, an operation with respect to the set of data can be continued by using the other one of the external information-storage devices.
In RAID-2, a set of data is stripped on an individual bit basis, and a Hamming code is produced for the stripped data as parity. The added error correction code is dispersed and stored in a plurality of external information-storage devices.
In RAID-3, a set of data is stripped on an individual bit or byte basis. Parity is used as an error correction code. The parity is stored in an external information-storage device that is assigned to exclusively store the error correction code.
In RAID-4, a set of data is stripped on a plurality of bytes basis. Parity is used as an error correction code. The parity is stored in an external information-storage device that is assigned to exclusively store the error correction code.
In RAID-5, a set of data is stripped on a plurality of bytes basis. Parity is used as an error correction code. Parities are dispersed and stored in a plurality of external information-storage devices.
Generally, RAID-0, RAID-1 and RAID-5 are used from among the above-mentioned RAID-0 to RAID-5, in consideration of a data transfer speed and an overhead necessary for various processes.
A detailed description will now be given of RAID-5, which is frequently used.
FIG. 2
is an illustration for explaining a process performed in RAID-5. FIG.
2
-(A) indicates a string of input data; and FIG.
2
-(B) indicates a plurality of data blocks produced by stripping the input data and storing the stripped input data in the plurality of external information-storage devices.
In RAID-5, a data block obtained by stripping a set of data on a several-byte basis and a corresponding error correction code are dispersed and stored in a plurality of external information-storage devices. For example, when the input data string A
1
to A
28
shown in FIG.
2
-(A) is input, a plurality of single data blocks are formed by dividing the input data string by each predetermined number of sets of individual data. Accordingly, the set of input data is divided into data blocks BL
1
to BL
7
.
The data blocks BL
1
to BL
7
are stored in a plurality of external information-storage devices
3
-
1
to
3
-
4
. At this time, an exclusive OR operation is performed with respect to the data blocks BL
1
to BL
7
so as to produce parities P1 to P3. The parities P1 to P3 are dispersed and stored in the external information-storage devices
3
-
1
to
3
-
4
so that parities are not concentrated in a particular one of the external information-storage devices.
When the set of data A
1
to A
28
is read, the data blocks BL
1
to BL
7
and the parities P1 to P3 are read concurrently so that the original set of data can be restored in a short time. That is, a data transfer speed can be increased by concurrently accessing the external information-storage devices
3
-
1
to
3
-
4
so as to concurrently read a plurality of data blocks and parities.
In order to establish the RAID system, one of a hardware method and a software method can be selected. When the hardware method is selected, the calculation of parities, the stripping of the set of data and the restoration of the original set of data are performed by an exclusive circuit such as a RAID controller. According to the hardware method, a high-speed processing can be achieved but a complex operation is required for introducing new hardware and such hardware is expensive. When the software method is selected, new hardware is not required and an installation cost is low. However, the software method has a lower performance than the hardware method.
The RAID system has, on one hand, an advantage that the improvements in data transfer speed and reliability can be achieved but, on the other hand, there is a disadvantage that a storage capacity and a method of RAID cannot be easily changed once the system is put into practical use. For example, in order to increase a storage capacity, it is required to perform formatting of recording media, backup of existing data and restoration of data. This causes a stop of service or a decrease in the data transfer speed during an operation for increasing the storage capacity.
FIG. 3
is a flowchart of an operation for changing a structure of a conventional RAID system.
When a structure of the RAID system is changed, data stored in all of the external information-storage devices of the RAID system is read in step S
1
-
1
, and the read data (backup data) is stored i
Karasudani Akira
Nakano Yasuhiko
Okada Yoshiyuki
Bragdon Reginald G.
Fujitsu Limited
Greer Burns & Crain Ltd.
Vital Pierre M.
LandOfFree
Information storage system for redistributing information to... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Information storage system for redistributing information to..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Information storage system for redistributing information to... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3154018