Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
1998-10-29
2001-04-24
Iqbal, Nadeem (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C711S114000
Reexamination Certificate
active
06223300
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a disk array apparatus comprising storage devices and, more particularly, to a disk array apparatus capable of changing a previously established array parameter in accordance with logs of past access to the storage devices so that data areas in the storage devices are reconstructed.
2. Description of the Related Art
There have been developed disk array apparatuses each connecting a plurality of hard disk drives (simply called disks hereunder) through a parallel interface to constitute a mass storage disk subsystem that may be accessed rapidly for enhanced storage device performance.
In such a disk array apparatus, data are stored in a distributed manner throughout the configured disks for parallel high-speed access. If a parity disk is provided, a failed disk may be remedied by resorting to parity information. That is, the faulty disk may be removed and replaced on a hot-line basis by a new disk, whereby the data held on the defective disk are recovered.
Where a spare disk is provided for use in an emergency, the data stored on a failed disk may be created anew on the spare disk. The faulty disk is then isolated from the system.
In the conventional disk array apparatus, the data area in each of the configured disks is managed in units of stripes so that the disks are accessed in parallel. More specifically, the data area is managed in units of segments corresponding to continuous address locations.
In that setup, as shown in
FIG. 7A
, a mass read or write instruction is processed at high speed through parallel access to a plurality of disks. Or as shown in
FIG. 7B
, separate read or write instructions from a plurality of applications are distributed to the suitable disks for high-speed parallel processing.
In conventional disk array apparatuses, once the disk subsystem is configured, array parameters such as disk stripe sizes are rarely altered from initially established value in keeping with use frequency or the like. It follows that the parameters set for access to the configured disks may not necessarily remain optimal.
The trends of access to the configured disks vary significantly depending on the application type, data sizes, the number of clients, use frequency and other factors. In fact, it is not very probable that the initially established array parameters such as the stripe size, disk count and RAID level would remain valid for an extended period of time. The efficiency of access to disks has often deteriorated during system operation.
Illustratively, it often happens that, as shown in
FIG. 8A
, a plurality of processing instructions of large access sizes cannot be processed in parallel and that, as depicted in
FIG. 8B
, requests of small access sizes concentrated on a specific disk may prevent parallel access thereto.
Despite those problems, few attempts have been made to alter the initially established array parameters such as the stripe size in conventional disk array apparatuses.
If the stripe size were to be changed, the host system of the disk array apparatus would have to be stopped for an appreciably long period of time to make backup copies of necessary data. With the old stripe size replaced by a new stripe size, the backup copies must be restored.
It would also be necessary to check the validity of the stripe size by retrieving access logs from the apparatus for manual analyses. The analyses can amount to a time-consuming, labor-intensive process. Sometimes there were no means to retrieve the access logs.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide a disk array apparatus which, without stopping its host system, allows an optimum array parameter to be determined in keeping with disk access trends so that data areas in configured storage devices such as disks are reconstructed accordingly.
Further objects, features and advantages of this invention will become more apparent upon a reading of the following description and appended drawings.
REFERENCES:
patent: 5408644 (1995-04-01), Schneider et al.
patent: 5511227 (1996-04-01), Jones
patent: 5559956 (1996-09-01), Sukegawa
patent: 5574851 (1996-11-01), Rathunde
patent: 5619690 (1997-04-01), Matsumani et al.
patent: 5724539 (1998-03-01), Riggle et al.
patent: 5765204 (1998-06-01), Bakke et al.
patent: 5809224 (1998-09-01), Schultz et al.
patent: 5950230 (1999-09-01), Islam et al.
patent: 5961652 (1999-10-01), Thompson
patent: 6058455 (2000-05-01), Islam et al.
patent: 6058489 (2000-05-01), Schultz et al.
patent: 8328759 (1996-12-01), None
patent: WO 97/07461 (1997-02-01), None
XP-002144050; Gerhard Weikum and Peter Zabback; Tuning of Striping Units in Disk-Array-Based File Systems; Department of Computer Science Information Systems-Databases; Feb. 2, 1992, pp. 80-86.
Alps Electric Co. ,Ltd.
Brinks Hofer Gilson & Lione
Iqbal Nadeem
LandOfFree
Disk array apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Disk array apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Disk array apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2499071