Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
2001-02-20
2004-11-16
Beausoliel, Robert (Department: 2113)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C714S005110, C714S006130, C714S006130, C714S006130, C714S042000
Reexamination Certificate
active
06820212
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATIONS.
Not Applicable
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT.
Not Applicable
Reference to a “Microfiche appendix.”
Not Applicable
BACKGROUND OF THE INVENTION.
1. Field of the Invention
This invention relates to RAID systems with provisions for maintaining the speed or channel capacity of the system under conditions of single component failure.
2. Description of Related Art Including Information Disclosed Under 37 CFR 1.97 AND 37 CFR 1.98.
The present invention is a RAID system which has redundant connections between active storage array controllers and the arrays of storage units they control, spare storage units in each array, and a passive storage array controller which assumes the control of the array of storage units of any failed storage array controller. Thus the failure of any one connector, storage unit, or storage array controller does not affect the channel capacity or speed of the RAID system of this invention.
U.S. Pat. No. 5,651,110 discloses a RAID system with two second level storage array controllers each of which control an array of disk drives. Each second level storage array controller is controlled by a separate first level storage array controller, which, in turn, communicates with the computer. In the event of a failure of a second level storage array controller, control of the array of disks assigned to the failed storage array controller is assumed by the intact second level storage array controller, which now controls both its original disks and the disks of the failed second level storage array controller. The channel capacity of the RAID system is thereby reduced by half under conditions of a failed second level storage array controller.
U.S. Pat. No. 5,787,070 discloses a global computer network packet switching system in which a number of active service modules are backed up by a normally passive redundancy module which takes the load when one of the active service modules fails. The communication system has no provisions for data storage.
U.S. Pat. No. 5,790,775 discloses a data storage system with a SCSI environment. The system involves two storage array controllers in dual-active, redundant configuration, and associated physical storage media. Failure of one storage array controller results in the other storage array controller assuming the control of all of the SCSI units (failover). The reverse operation, wherein the defective storage array controller is repaired or replaced and assumes control of its portion of the storage media, is termed “failback”. The channel capacity of the data storage system is reduced by half under conditions of a failed storage array controller.
U.S. Pat. No. 5,848,230 discloses a RAID system in which there is triple replication of all subsystems. It has three storage array controllers, one active and two which are normally passive and are used only in case of the failure of the active storage array controller and (subsequently) the secondary storage array controller. In addition, triplicate subsystems such as cooling and power subsystems are included. This system provides highly reliable and continuous availability of storage service and an undiminished channel capacity. The provision of two normally passive storage array controllers for each active storage array controller is a major contributor to the cost of this system.
U.S. Pat. No. 5,872,906 discloses a RAID system with provisions for allocating a spare disk unit in case of a disk failure. It includes two substorage array controllers which are provided for the common buses thereby distributing the processing functions of the storage array controllers and reducing a load. No provisions for failure of a storage array controller are disclosed.
U.S. Pat. No. 5,922,077 discloses a RAID system with two storage array controllers and a fail-over switch which routes the data from the storage array controller of a failed communication path to the operating storage array controller, which then handles the load of both storage array controllers. The channel capacity is reduced when one storage array controller is handling both loads.
U.S. Pat. No. 5,944,838 discloses a RAID system with a redundant storage control module (RDAC) in which two queues of pending I/O requests are maintained for a single array of storage devices. The redundant queue takes over on the failure of the active queue. The redundant queue copies each I/O request sent to the active path which minimizes the time required for the redundant queue to take over the functions of the active queue.
U.S. Pat. No. 6,073,218 discloses an apparatus for coordinating multiple RAID storage array controllers' access to a single array of storage devices. Each of a number of storage array controllers process different I/O requests on an array of common shared storage devices. One storage array controller is designated primary with respect to the storage devices. Concurrent access to the storage devices is coordinated by the storage array controllers.
None of the prior art RAID systems achieves the advantages of the present invention, that of preserving the channel capacity or speed of the system in the face of failure of a connector, storage unit, or storage array storage array controller, all with minimal redundancy of components and minimal cost. In particular, in this invention a single passive storage array controller is available to replace any one of two or more active storage array controllers when an active storage array controller becomes defective.
BRIEF SUMMARY OF THE INVENTION
This invention is RAID system which is able to function with undiminished channel capacity or speed despite the failure of any one component, with minimal redundancy of components. The system comprises n active storage array controllers, n arrays of storage units, each active storage array controller controlling one or more arrays of storage units, and one only passive storage array controller. The passive storage array controller is connected by two connectors to each active storage array controller, and is able to control any one of the arrays of storage devices. In the event of failure of any of the active storage array controllers, the passive storage array controller assumes the identity of the failed storage array controller and assumes control of the array of storage devices of the failed storage array controller. Since each array of storage units contains a spare unit which becomes active when one storage unit in the array fails, the RAID system of this invention is able to function with undiminished capacity in the event of failure of any one storage unit or any one storage array controller. In another embodiment, each storage unit is connected to controllers by two connectors, and this embodiment is, in addition, able to function in the event of failure of any one connector.
The objective of this invention is to provide a RAID system with undiminished capacity despite the failure of any one storage unit, any one connector, or any one storage array controller.
Another objective is to provide a RAID system which produces a signal for the operator in the event of failure of any component.
Another objective is to provide a RAID system which automatically substitutes a replacement for a failed storage unit or storage array controller.
Another objective is to provide a RAID system with redundant connectors connecting the storage units and the storage array controllers.
Another objective is to provide a RAID system with several active storage array controllers which normally control the arrays of storage units and with one passive storage array controller which assumes control the storage units of any active storage array controller which fails.
Another objective is to provide a RAID system capable of functioning with undiminished channel capacity in the event of failure of a connector, storage unit, or storage array controller with minimal redundancy of components.
Another objective is to provide a RAID system capable of functioning with undiminished channel capacity i
Duchesne Raymond
Rauscher Tomlinson
Beausoliel Robert
Digi-Data Corporation
Puente Emerson
Ramsey William S.
LandOfFree
RAID system having channel capacity unaffected by any single... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with RAID system having channel capacity unaffected by any single..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and RAID system having channel capacity unaffected by any single... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3355397