JP2005284449A - Information recording medium array system and its restoration method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a disk array system for efficiently restoring a disk array when the disk array is put into an inoperable state. <P>SOLUTION: A disk array controller 13 is provided with an inoperable disk recording table 1 for recording an order in which a hard disk device is put into an inoperable state and the identification information and failure factor of the hard disk device. When the disk array is put into an inoperable state, a writing processing part 2 records the identification information of the hard disk device which is put into the inoperable state and its order and failure factor in the inoperable disk recording table 1. At the time of restarting the hard disk device after removing the factor of the inoperable state, a decision processing part 3 refers to the inoperable disk recording table 1, and tests any hard disk device other than the hard disk device which has initially been put into the inoperable state, and a reconfiguration processing part reconfigures the hard disk device which is judged to be available by the testing means as a disk array. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an information recording medium array system having a medium array composed of information recording media such as a plurality of hard disk devices connected to a computer apparatus, and a medium array control apparatus for controlling the medium array, and a recovery method thereof. In particular, an information recording medium array system configured to be able to recover even when a failure occurs in two or more information recording media and the medium array becomes inoperable, and a recovery method therefor It is about.

  In general, a disk array constitutes one storage system by a plurality of hard disk devices. Even if an abnormality occurs in one of the devices, the storage contents of the device in which the abnormality has occurred are restored based on the storage contents of the other hard disk devices, and the operation is continued. As a typical example of such a disk array system, there is a disk array system configured with redundant RAID (redundant array of independent disks). RAID-1, 3, 5, etc. are often used as redundant RAID.

  For example, an outline of a conventional disk array system described in Patent Document 1 will be described. When the system is powered on, the connection state of the disk array is checked, and the information is stored in the configuration information file of the auxiliary recording device. Next, data is read and written using a plurality of disk devices as the operation of the disk array system. During the free time, it is checked whether each disk device has an abnormality. If there is no abnormality, the process is returned and the check is repeated.

  If an error is detected as a result of the check, check the free capacity of all the disk devices other than the abnormal disk device and the data capacity stored in the disk device where the error occurred. If the difference is less than a predetermined capacity, the disk array is output as being abnormal, and the system is stopped.

  If the difference in capacity is within a predetermined range, the data stored in the failed disk unit is restored using the data stored in the normal disk unit, and the data is restored to normal. Data distribution processing is performed by dividing and storing the data in a different disk device.

Also, a disk array system that appropriately performs data recovery is known from Patent Document 2. According to the disk array system described in Patent Document 2, even when an abnormality occurs in two of the plurality of disk devices constituting the disk array, the occurrence of the abnormality is managed using a flag, and all the disk devices are managed. By permitting reading, the stored contents can be read from a disk device in which no abnormality has occurred.
Japanese Patent Laid-Open No. 08-190461 JP-A-11-184642

  In the conventional disk array system as described above, the first problem is that it becomes difficult for the disk array control device to maintain redundancy when a plurality of hard disk devices are in a failure state at the same time. In other words, the disk array control device checks the normality of the operation based on the parity of the information stored in a distributed manner in the disk array configuration to maintain redundancy. However, when a plurality of hard disk devices are in a failure state, it is difficult to calculate normal information from the parity.

  A second problem is that recovery is impossible when a plurality of hard disk devices are recognized as being in a failure state. If one hard disk device becomes inoperable, the hard disk device that has become inoperable is disconnected from the disk array configuration (RAID configuration), and the disk array operation is continued with the remaining hard disk devices. In this case, the data of the disconnected hard disk device and the remaining hard disk devices are not synchronized. Further, when the next hard disk device becomes inoperable, the hard disk device is also disconnected from the disk array configuration, so that a so-called RAID operation in which a plurality of hard disk devices are operated as one disk device is disabled. Become.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a disk array system including a disk array composed of a plurality of hard disk devices. It is an object of the present invention to provide a disk array control device that can be recovered even when it becomes inoperable.

In order to achieve the above object, according to the present invention, in an information recording medium array system comprising a medium array composed of a plurality of information recording media and a medium array controller for controlling the medium array, the medium array The control device includes an inoperable medium recording table having fields for recording the order in which the plurality of information recording media have become inoperable and identification information of the information recording medium, and the medium array in an inoperable state. Writing means for writing the identification information and the order of the information recording media that have become inoperable in the inoperable medium recording table, and the inoperable medium recording table at the time of restart after removing the cause of the inoperable state. An inspection means for inspecting an information recording medium other than the information recording medium which is first inoperable with reference to the inspection means; Information recording medium array system for the recovery means for reconstructing the information recording medium can be judged that allows use as a medium array, characterized in that is provided, is provided.
In order to achieve the above object, according to the present invention, in an information recording medium array system having a medium array composed of a plurality of information recording media, information is obtained when a plurality of information recording media are in a failure state. This is a recovery method in the case where the recording medium array system becomes inoperable. Except for the information recording medium that has failed first, the information recording medium that has failed second and later is inspected and reused. In the process of verifying whether or not it is possible, and when it is determined that any of the information recording media in the second or later failure state is reusable, the information recording in the first failure state And a method of recovering an information recording medium array system, comprising the step of configuring a medium array excluding the medium and forcibly starting the newly configured medium array in an online state. It is.
Preferably, the medium array has redundancy in information storage. Preferably, the information recording medium is a hard disk device.

  In the medium array system of the present invention, an inoperable medium having fields for recording the order in which a plurality of information recording media become inoperable, identification information of the information recording medium, and the cause of failure in the medium array control device. A recording table, writing means for writing to the inoperable medium recording table, inspection means for inspecting the information recording medium that has become inoperable, and recovery means are provided. The writing means writes the identification information and the order of the information recording media that have become inoperable in the inoperable medium recording table when the disk array becomes inoperable, so that the inspection means eliminates the cause of the inoperable state. An information recording medium in which a medium other than the information recording medium that was initially inoperable is inspected by referring to the inoperable disk recording table at a subsequent restart, and the recovery means can determine that the inspection means can use the medium As a media array.

  As a result, even if the medium array becomes inoperable, it can be used except for the information recording medium that has become inoperable at the beginning of the cause of the direct cause of the inoperable state. By checking the possible information recording media and forcibly starting up the remaining information recording media that have been determined to be usable online, the media array can be reconfigured with the available information recording media. The media array operation can be continued.

  According to the information recording medium array system and the restoration method of the present invention, the redundancy of the disk array system can be further increased. The majority of data loss is a case where a plurality of information recording media fall into a failure state at the same time.According to the present invention, it is remedied that a plurality of information recording media become inoperable at the same time. Can be increased. For this reason, the method of the information recording medium array system of the present invention can be suitably used not only as a hard disk array system but also as a recovery means for other large-capacity storage devices.

  Next, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing a system configuration of a main part of a disk array system 100 according to the present invention. As shown in FIG. 1, the disk array system 100 includes a disk array 5 and a disk array controller 13 that controls the disk array 5. In the disk array controller 13, 1 is an inoperable disk recording table, 2 is a fail order write processing unit as a recording unit, 3 is a reuse determination processing unit as an inspection unit, and 4 is a RAID configuration reconfiguration processing as a recovery unit. Part. Reference numeral 5 denotes a disk array, and reference numeral 6 denotes a configuration information table in which state information of the disk array is recorded.

  7 is an ID field in which ID information for identifying a plurality of hard disk devices constituting the disk array is recorded; 8 is a status field in which status information of the plurality of hard disk devices constituting the disk array is recorded; 9 is an order field in which information on the order of hard disk devices in an inoperable state (fail state) is recorded, 10 is an ID field in which ID information for identifying hard disk devices in an inoperable state is recorded, and 11 is an inoperable state A failure cause field 12 in which information on the cause of failure of the hard disk device is recorded, and 12 is a fail information operation unit. The disk array 5 is composed of a plurality of hard disk devices 14.

  Referring to the figure, a disk array control device 13 for controlling a disk array 5 composed of a plurality of hard disk devices 14 is connected to a computer device (not shown), and an input / output request for data processing from the computer device. The data is written to and read from the hard disk device.

  In this case, the disk array control device 13 manages a plurality of hard disk devices 14 of the disk array 5 configured in RAID according to the status information and ID information recorded in the configuration information table 6, and 5 is controlled to perform RAID operation. Although the RAID operation of the disk array 5 has various operation modes, it is publicly known and is not directly related to the contents of the present invention, so a detailed description is omitted. However, in the redundant RAID operation, all hard disks are used. When the device 14 is operating normally, it is registered in the status field 8 of the configuration information table 6 that all the hard disk devices 14 are operating online. If one hard disk device 14 becomes inoperable, the hard disk device is disconnected from the disk array 5 by the RAID operation, and the operation is continued by the remaining hard disk devices. Then, the status in the configuration information table 6 of the hard disk device that has been disconnected due to the occurrence of a failure is rewritten as “OFFLINE”.

  When such a disk array 5 is controlled and operated in RAID, when two or more hard disk devices fail and the system becomes inoperable (fail state), the fail information operation unit 12 is activated. Then, the fail order write processing unit 2, the reuse determination processing unit 3, and the RAID configuration reconfiguration processing unit 4 perform processing in sequence to restore the RAID operation in the disk array 5. The inoperable disk recording table 1 is used for this recovery operation. The inoperable disk recording table 1 includes fields for recording the order in which the plurality of hard disk devices 14 constituting the disk array 5 are in an inoperable state, identification information of each hard disk device 14 and the cause of failure. Is provided. The recovery operation is performed based on the information recorded in each of these fields.

  That is, when the disk array 5 is in a failed state, the fail order writing processing unit 2 records the identification information, the order, and the cause of failure of the failed hard disk device in the inoperable disk recording table 1. To do. Next, after the cause of the failure state is removed, the reuse determination processing unit 3 refers to the inoperable disk recording table 1 after eliminating the freeze state, for example, other than the hard disk device that has first failed. The hard disk devices 14 are inspected based on failure cause information to determine whether or not these hard disk devices 14 can be reused. For example, it is determined whether or not the cause of the failure of the hard disk device is a problem of the hard disk device itself. In other words, this is the case where the direct cause of the failure is in the hard disk device that has become inoperable for the first time, and the hard disk device that has failed for the second time or later is indirectly in the failed state. This is because there are many. The RAID configuration reconfiguration processing unit 4 forcibly starts up the hard disk device 14 that has been determined to be usable by inspection at the time of restart, and reconfigures it as the disk array 5. That is, the configuration information table 6 is rewritten. As a result, the RAID operation is continued.

  As described above, the disk array control apparatus according to the present embodiment handles the recording area and the inoperability state information for recording the order and cause of the inoperability of the hard disk devices in the redundant RAID configuration. When a processing unit is added and there are multiple hard disk devices that are inoperable, the hard disk devices other than the hard disk device that initially became inoperable when the system is restarted after removing the cause are inspected. If it can be determined, for example, if it is a transient failure, it is forcibly launched in an online state.

  In a redundant RAID operation, even if one hard disk device becomes inoperable, the hard disk device that has become inoperable is disconnected from the RAID configuration, and RAID operations are performed on the remaining hard disk devices. However, if the RAID operation is continued as it is, the data of the disconnected hard disk device and the remaining hard disk devices are not synchronized. Further, when another hard disk device becomes inoperable, the hard disk device is also disconnected from the RAID configuration, and the RAID operation becomes impossible.

  In this case, since the access to the RAID operation is stopped, the data of the hard disk device that has become inoperable later and the data of the remaining hard disk devices are kept in synchronization, so the cause of the inoperable state is a temporary failure. If the cause is other than the hard disk device, the hard disk device can be used as it is if the cause is removed, and the second and subsequent hard disk devices in which data synchronization is maintained are re-installed in the RAID. The RAID can be reproduced by (online state).

  According to the disk array control apparatus of the present invention, it is possible to further increase the redundancy of the conventional disk array apparatus. Since most of the data loss is a case where a plurality of hard disk devices fall into a failure state at the same time, the robustness of information can be enhanced by applying the present invention. The method of the disk array control device is also preferably used as a recovery means for other mass storage devices.

1 is a block diagram showing a system configuration of a main part of a disk array system according to the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Inoperable disk recording table 2 Fail order writing processing unit 3 Reuse determination processing unit 4 RAID configuration reconfiguration processing unit 5 Disk array 6 Configuration information table 7 ID field 8 Status field 9 Order field 10 ID field 11 Failure cause field 12 Fail Information operation unit 13 Disk array controller 14 Hard disk device 100 Disk array system

Claims (6)

In an information recording medium array system having a medium array composed of a plurality of information recording media and a medium array control device for controlling the medium array,
The media array controller includes
An inoperable medium recording table having fields for recording the order in which the plurality of information recording media are inoperable and identification information of the information recording medium;
A writing means for writing identification information of an information recording medium that has become inoperable and an order thereof in the inoperable medium recording table when the medium array becomes inoperable;
An inspection means for inspecting an information recording medium other than the information recording medium first inoperable by referring to the inoperable medium recording table at the time of restart after the cause of the inoperable state is removed;
A recovery means for reconfiguring the information recording medium determined to be usable by the inspection means as a medium array;
An information recording medium array system comprising: 2. The information recording medium array system according to claim 1, wherein the medium array has redundancy in information storage. The inoperable medium recording table is also provided with a field for recording the cause of failure of the information recording medium that has become inoperable, and the failure of the information recording medium in which the writing means has become inoperable. 3. The information recording medium array system according to claim 1, wherein the cause is written in a corresponding field of the failure cause field. 4. The information recording medium array system according to claim 1, wherein the information recording medium is a hard disk device. In an information recording medium array system having a medium array composed of a plurality of information recording media, this is a recovery method when the information recording medium array system becomes inoperable due to a plurality of information recording media having failed. And
Excluding the information recording medium that has failed first, the process of inspecting the information recording medium that has failed second and later, and verifying whether it can be reused, and the second and subsequent failures If it is determined that any of the recorded information recording media can be reused, the media array is configured except for the information recording medium that has first failed, and the newly configured media array is forced. And a method for recovering an information recording medium array system, comprising: 6. The information recording medium array system recovery method according to claim 5, wherein a redundant RAID operation is performed in the medium array.

Images