JP2023134170A - Storage medium management device, method for managing storage medium, and storage medium management program - Google Patents

Abstract

To precisely predict a storage medium with a high failure risk and reduce a rebuilding time.SOLUTION: An HDD information collecting unit 211 collects the operational state of a plurality of physical disks HDD1 to HDD3 forming an RAID array. A statics information analysis unit 212 determines whether there is a relative difference in the operational state of the collected physical disks HDD1 to HDD3. A failure HDD prediction unit 213 predicts a physical disk with the highest failure risk on the basis of the relative difference in the operational state. An HDD copy control unit 214 copies the data of the physical disk predicted to have the highest failure risk into a standby disk SHDD.SELECTED DRAWING: Figure 1

Description

The present invention relates to a storage medium management device, a storage medium management method, and a storage medium management program.

In recent years, with the increase in the capacity of physical disks in RAID (Redundant Arrays of Inexpensive Disks, or Redundant Arrays of Independent Disks) systems, when a physical disk fails, the RAID rebuild process requires a large amount of time. ing. Therefore, in order to shorten the rebuild processing time, for example, when the error statistics value exceeds a predetermined threshold, the address range of the suspect disk device is set to A technique has been proposed in which data is sequentially copied to a spare disk and restored while specifying the data (for example, Patent Document 1). The rebuild process is performed by constructing and operating a RAID system using physical disks such as multiple HDDs, and if one disk fails in that RAID system, the failure will be detected from other HDDs that are not failed. Processing such as restoring and reconstructing data on a disk.

Japanese Patent Application Publication No. 2006-139339

However, in the method disclosed in Patent Document 1, additional values are set in advance for various types of failures in disk devices (storage media), such as motor stop, medium defect errors, mode errors, etc., and countermeasures are taken for each error occurrence. The added value is defined as the error statistics added value, and a disk device for which the error statistics added value exceeds a predetermined threshold is identified as a suspect disk device. However, just because the error statistics added value exceeds a predetermined threshold value does not necessarily mean that the suspected disk device will fail in the future. Therefore, it is not possible to accurately identify a disk device with a high risk of failure in the future, which may cause other disk devices to fail, resulting in the problem of requiring a large amount of time for the rebuild process. .

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a storage medium management device, a storage medium management method, and a storage medium management program that more accurately predict storage media with a high risk of failure and shorten rebuild time.

In order to solve the above-mentioned problems, one aspect of the present invention includes a collection unit that collects the operating states of a plurality of storage media that constitute a RAID array, and a collection unit that collects the operating states of the plurality of storage media that are collected by the collection unit. an operating state determining means for determining whether there is a relative difference between the operating state determining means; a predicting means for predicting a storage medium with the highest failure risk based on a determination result by the operating state determining means; The present invention is characterized by comprising a copy control means for copying data of the storage medium having the highest failure risk to a standby storage medium.

Further, one aspect of the present invention is a storage medium management method for managing a plurality of storage media constituting a RAID array, the method comprising: collecting operational states of the plurality of storage media; determining whether there is a relative difference in the operating states of storage media; predicting a storage medium with the highest failure risk based on the result of the determination; and determining whether the predicted failure risk is copying data on the highest storage medium to a standby storage medium.

Further, one aspect of the present invention includes the step of collecting the operating states of the plurality of storage media in a computer of a storage medium management device that manages the plurality of storage media constituting the RAID array; determining whether there is a relative difference in the operating states of storage media; predicting a storage medium with the highest failure risk based on the result of the determination; and determining whether the predicted failure risk is The present invention is characterized by causing a step of copying data of the highest storage medium to a standby storage medium to be executed.

As described above, it is possible to more accurately predict storage media with a high risk of failure and to reduce the rebuild time.

1 is a block diagram showing the configuration of a server system to which a RAID controller (storage medium management device) 20 according to the present embodiment of the present invention is applied. 3 is a flowchart for explaining operations periodically executed by the RAID controller 20 according to the present embodiment. 3 is a flowchart for explaining the normal operation of the RAID controller 20 according to the present embodiment. FIG. 3 is a schematic diagram showing a disk control operation by the RAID controller 20 according to the present embodiment. FIG. 3 is a schematic diagram showing a disk control operation by the RAID controller 20 according to the present embodiment. FIG. 3 is a schematic diagram showing a disk control operation by the RAID controller 20 according to the present embodiment. FIG. 3 is a schematic diagram showing a disk control operation by the RAID controller 20 according to the present embodiment. FIG. 3 is a schematic diagram showing a disk control operation by the RAID controller 20 according to the present embodiment. FIG. 3 is a schematic diagram showing a disk control operation by the RAID controller 20 according to the present embodiment. FIG. 3 is a schematic diagram showing a disk control operation by the RAID controller 20 according to the present embodiment. FIG. 1 is a block diagram showing the minimum configuration of a storage medium management device according to the present embodiment.

Embodiments of the present invention will be described below with reference to the drawings.

A. Embodiment FIG. 1 is a block diagram showing the configuration of a server system to which a RAID controller (storage medium management device) according to the present embodiment is applied. The server system 1 includes an OS (Operating System) 10, a RAID controller 20, and a disk array 30. The OS 10 includes a RAID driver 11 and a RAID management utility 12. The RAID driver 11 is software for transmitting and receiving (reading/writing) data to and from the disk array 30. The RAID management utility 12 is software used to manage reading/writing of data to/from the disk array 30 at the OS level.

The RAID controller 20 includes RAID firmware 21, and the RAID firmware 21 includes an HDD information collection section 211, a statistical information analysis section 212, a failed HDD prediction section 213, and an HDD copy control section 214. The disk array 30 includes physical disks HDD1, HDD2, HDD3 and a standby disk SHDD that configure RAID5. RAID5 is a technology that manages multiple external storage devices (hard disks, etc.) as a single device. In this embodiment, the physical disks HDD1, HDD2, HDD3, and standby disk SHDD are assumed to be hard disks, but are not limited to this, and may be semiconductor storage media such as SSD (Solid State Drive).

The HDD information collection unit 211 periodically collects statistical information of each physical disk HDD1 to HDD3 of the disk array 30. The statistical information is information indicating at least one operating state such as operating time, response speed, error rate, and number of registered alternative sectors. The statistical information analysis unit 212 analyzes the collected statistical information including at least one operating state of the physical disks HDD1 to HDD3, such as operating time, response speed, error rate, and number of registered alternative sectors. The analysis method will be described later.

The failure HDD prediction unit 213 determines which of the plurality of physical disks HDD 1 to HDD 3 connected to the RAID controller 20 will be selected in the future based on the analysis results of the statistical information of the physical disks HDD 1 to HDD 3 analyzed by the statistical information analysis unit 212. Identify the HDDm (m=1, 2, or 3) with the highest risk of failure. The HDD copy control unit 214 copies the data of the identified HDDm with the highest risk of failure in the future to the standby disk SHDD.

B. Operations of the Embodiment FIG. 2 is a flowchart for explaining operations periodically executed by the RAID controller 20 according to the present embodiment. 4 to 6 are schematic diagrams showing disk control operations by the RAID controller 20 according to this embodiment.

The flowchart shown in FIG. 2 is executed periodically (for example, every week). The HDD information collection unit 211 collects statistical information such as the operating time, response speed, error rate, and number of registered alternative sectors of each physical disk HDD1 to HDD3 of the disk array 30 (step S10). The statistical information analysis unit 212 analyzes the collected statistical information indicating the operating time, response speed, error rate, and number of registered alternative sectors of the physical disks HDD1 to HDD3, and determines whether there is a risk of future failure in the operating status of the physical disks HDD1 to HDD3. It is determined whether there is a difference sufficient to predict something (step S12). For example, by analyzing the statistical information of the physical disks HDD1 to HDD3 of the members that make up the logical drive, we will analyze the statistical information of the physical disks HDD1 to HDD3 of the members that make up the logical drive. It is determined whether there is a relative difference (difference indicating deterioration) in at least one operating state.

Then, if there is no difference in at least one operating state included in the collected statistical information of each of the physical disks HDD1 to HDD3 (NO in step S14), the HDD copy control unit 214 selects an arbitrary physical hard disk HDDn(n =1, 2, or3) (for example, HDD2), and copies the data of the physical hard disk HDDn to the standby disk SHDD (step S16). For example, as shown in FIG. 4, the physical disk HDD2 is selected and the data on the physical disk HDD2 is copied to the standby disk SHDD. After that, the process ends.

On the other hand, if there is a difference in at least one operating state included in the collected statistical information of each of the physical disks HDD1 to HDD3 that is sufficient to predict that there is a risk of failure in the future (YES in step S14), ), the failure HDD prediction unit 213 calculates the analysis results analyzed by the statistical information analysis unit 212, that is, at least one operating state of the physical disks HDD1 to HDD3 (operating time, response speed, error rate, number of registered alternative sectors, etc.). Based on the relative difference between the physical disks HDD1 to HDD3 connected to the RAID controller 20, the physical disk HDDm (m=1, 2, or3) with the highest risk of failure in the future is identified (step S18).

For example, the failure HDD prediction unit 213 determines whether any of the member physical disks HDD1 to HDD3 constituting the logical drive have a long operating time, a slow response speed, a large error rate, a large number of alternative sectors, etc. A physical hard disk having at least one relative difference in operating state with respect to the physical disk of is identified as the physical disk HDDm with the highest risk of failure. For example, as shown in FIG. 5, if physical disk HDD2 has a slightly longer operating time, slower response speed, or a larger number of alternative sectors registered than other physical disks, physical disk HDD2 may , is identified as the physical disk HDDm that is predicted to have the highest risk of failure in the future.

Note that for "slight", a threshold value may be set in advance for the "difference", and when the threshold value is exceeded, the physical disk may be predicted to be the physical disk HDDm with the highest risk of failure in the future. Further, as to which operating state to use in the determination of operating time, response speed, and number of alternative sector registrations, each operating state may be weighted or a priority may be set. Furthermore, at least one of the operating time, response speed, and number of alternative sector registrations may be used as a judgment condition, or two or more may be used as a judgment condition in combination.

Next, the failure HDD prediction unit 213 determines that the currently identified physical disk HDDm with a high failure risk has a higher failure risk than the previously identified physical disk HDDp with a high failure risk (p=1, 2, or 3). It is determined whether or not (step S20). Whether the failure risk is high or low can be determined based on the respective operating time, response speed, number of registered alternative sectors, etc. If the failure risk of the currently identified physical disk HDDm with a high failure risk is lower than the previously identified physical disk HDDp with a high failure risk (NO in step S20), the process ends without doing anything. do.

On the other hand, if the currently identified physical disk HDDm with a high failure risk has a higher failure risk than the previously identified physical disk HDDp with a high failure risk (YES in step S20), the HDD copy control unit 214 The data of the identified physical disk HDDm with the highest risk of failure is copied to the standby disk SHDD (step S22).

For example, if physical disk HDD3 is identified as a physical disk HDDm with a high failure risk this time, the failure risk of the physical disk HDD3 is higher than that previously identified due to long operating time, slow response speed, large number of registered alternative sectors, etc. If the failure risk is higher than that of the physical disk HDD2, as shown in Figure 6, the physical disk HDD2 is switched to the physical disk HDD3, and the data of the physical disk HDD3, which has the highest failure risk identified this time, is transferred to the standby disk SHDD. make a copy. After that, the process ends.

FIG. 3 is a flowchart for explaining the normal operation of the RAID controller 20 according to this embodiment. Further, FIGS. 7 to 10 are schematic diagrams showing disk control operations by the RAID controller 20 according to this embodiment.

The flowchart shown in FIG. 3 is executed when a write request is received from the host OS 10, an application, or the like. In the RAID controller 20, when a write request is received from the host OS 10, an application, etc., the HDD copy control unit 214 writes data to the physical disk HDDm identified as having a high risk of failure, and also writes data to the standby disk SHDD. The same data is written (step S30). For example, if the physical disk with the highest risk of failure is the physical disk HDD2, as shown in FIG. 7, data is written to the physical disk HDD2 and the same data is also written to the standby disk SHDD.

Next, the RAID controller 20 determines whether the physical disk HDDm identified as having a high risk of failure has actually failed (step S32). If the physical disk HDDm identified as having a high risk of failure has not yet failed (NO in step S32), the process is ended without doing anything.

On the other hand, if the physical disk HDDm identified as having a high risk of failure actually fails (YES in step S32), the RAID controller 20 disconnects the failed physical disk HDDm and uses the same physical disk HDDm as the failed physical disk HDDm. The standby disk SHDD on which data was written is newly incorporated into the hard disk device (RAID) (step S34). For example, as shown in FIG. 8, if the physical disk HDD2 that has been identified as having a high risk of failure actually fails, the failed physical disk HDD2 is disconnected and the standby disk that was writing the same data as the failed physical disk HDD2 is The disk SHDD is incorporated into the RAID as the physical disk HDD2. Therefore, rebuilding work is no longer necessary, which prevents system failure and data loss. After that, the process ends.

Note that if the physical disk HDD 2 that is predicted to have a high failure risk by the failure HDD prediction unit 213 fails, but another physical disk, for example, the physical disk HDD 1, fails, as shown in FIG. 9, the remaining physical disk HDD 2 Then, an operation of writing data from the physical disk HDD3 to the standby disk SHDD, a so-called rebuild process, is performed. In this case, rebuild processing is required.

However, as described above, in this embodiment, if there is at least one operating state with respect to other physical disks, such as a long operating time, slow response speed, large error rate, or large number of registered alternative sectors, etc. A physical hard disk with a relative difference is identified as the physical disk HDDm with the highest risk of failure. Therefore, according to the present embodiment, compared to a method in which the determination criterion is simply that a uniquely set threshold value has been exceeded, it is possible to more accurately predict the physical disk HDDm that has the highest risk of failure in the future. Situations that require processing are unlikely to occur.

When the failed physical disk HDD2 is replaced with a new physical disk HDD2new, the physical disk HDD2new is automatically set as the standby disk SHDD, as shown in FIG. After completing the setting of the standby disk SHDD, the RAID controller 20 executes the flowchart shown in FIG. ) the data is copied to.

According to the embodiment described above, a physical hard disk that has at least one relative difference in operating state from other physical disks, such as a long operating time, a slow response speed, or a large number of registered alternative sectors, is selected. Since the physical disk HDDm with the highest risk of failure is identified as the physical disk HDDm with the highest risk of failure, it is possible to more accurately predict the physical disk with the highest risk of failure in the future. By copying the data on the physical disk HDDm, which has the highest risk of failure, to the standby disk SHDD before the failure occurs, even if the physical disk actually fails, it can be switched to the failed physical disk. Since the standby disk SHDD is incorporated into the RAID, the rebuild time can be shortened.

FIG. 11 is a block diagram showing the minimum configuration of the storage medium management device according to this embodiment.
The storage medium management device 50 according to the present embodiment includes at least a plurality of storage media 52 to 54 constituting a RAID array 51, a standby storage medium 55, a collection means 56, an operating state determination means 57, a prediction means 58, and a copy control means 59. All you have to do is prepare. The plurality of storage media 52 to 54 constitute a RAID array (RAID5). The collecting means 56 collects the operating states of the plurality of storage media 52-54. The operating state determining means 57 determines whether there is a relative difference in the operating states of the plurality of collected storage media 52-54. The prediction means 58 predicts the storage medium with the highest failure risk based on the determination result by the operating state determination means 57. The copy control means 59 copies the data of the storage medium with the highest failure risk predicted by the prediction means 58 to the standby storage medium 55.

Note that the program for realizing the functions of the processing unit in the present invention is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed, thereby controlling the benefit information. Processing may be performed. Note that the "computer system" herein includes hardware such as an OS and peripheral devices. Further, a "computer system" may include a plurality of computer devices connected via a network including the Internet, a WAN, a LAN, a communication line such as a dedicated line, etc. Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, magneto-optical disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into computer systems. Furthermore, a ``computer-readable recording medium'' refers to a storage medium that retains a program for a certain period of time, such as volatile memory (RAM) inside a computer system that serves as a server or client when a program is transmitted via a network. This shall also include things. Moreover, the above-mentioned program may be for realizing a part of the above-mentioned functions. Furthermore, it may be a so-called difference file (difference program) that can realize the above-mentioned functions in combination with a program already recorded in the computer system.

Furthermore, some or all of the functions described above may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each of the above-mentioned functions may be implemented as an individual processor, or some or all of them may be integrated into a processor. Further, the method of circuit integration is not limited to LSI, but may be implemented using a dedicated circuit or a general-purpose processor. Further, if an integrated circuit technology that replaces LSI emerges due to advances in semiconductor technology, an integrated circuit based on this technology may be used.

1 Server system 10 OS
11 RAID driver 12 RAID management utility 20 RAID controller 21 RAID firmware 211 HDD information collection unit 212 Statistical information analysis unit 213 Failure HDD prediction unit 214 HDD copy control unit 30 Disk array HDD1 to HDD3 Physical disk SHDD Standby disk

Claims (7)

a collection means for collecting operational states of a plurality of storage media constituting the RAID array;
operating state determining means for determining whether there is a relative difference in the operating states of the plurality of storage media collected by the collecting means;
prediction means for predicting a storage medium with the highest failure risk based on the determination result by the operating state determination means;
Copy control means for copying data of the storage medium with the highest failure risk predicted by the prediction means to a standby storage medium;
A storage medium management device comprising: The storage medium according to claim 1, wherein the copy control means writes data to the predicted storage medium and also writes the same data to the standby storage medium when there is a request to write data. Management device. The copy control means is characterized in that when the storage medium with the highest failure risk fails, the standby storage medium is incorporated as one of the plurality of storage media in place of the storage medium with the highest failure risk. 3. The storage medium management device according to claim 2. The copy control means is characterized in that, when it is determined by the operation state determination means that there is no difference in the operation states, the copy control means copies the data of any storage medium among the plurality of storage media to the standby storage medium. 4. The storage medium management device according to claim 3. 5. The operating state includes at least one of the operating time, response speed, error rate, and number of registered alternative sectors of each of the plurality of storage media. The storage medium management device described. A storage medium management method for managing a plurality of storage media constituting a RAID array, the method comprising:
collecting operational states of the plurality of storage media;
determining whether there is a relative difference in the operating states of the plurality of collected storage media;
predicting a storage medium with the highest failure risk based on the result of the determination;
Copying data of the storage medium with the predicted highest failure risk to a standby storage medium;
A storage medium management method characterized by comprising: A computer of a storage medium management device that manages a plurality of storage media constituting a RAID array,
collecting operational states of the plurality of storage media;
determining whether there is a relative difference in the operating states of the plurality of collected storage media;
predicting a storage medium with the highest failure risk based on the result of the determination;
Copying data of the storage medium with the predicted highest failure risk to a standby storage medium;
A storage medium management program that executes.

Images