JPH08147112A - Error recovery device for disk array device - Google Patents

Abstract

PURPOSE: To efficiently perform the recovery work by automatically performing the recovery processing without requiring hands neither replacement of a disk device by medium initialization of the disk device where a fault occurs. CONSTITUTION: If the frequency in error occurrence of one of disk devices 50 to 57 for data storage and a disk device 58 for redundant information storage in a disk array 5 exceeds a prescribed value, data of the disk device where error occurs is restored into an auxiliary disk device 59 by a first data restoration part 46; and when the restoration operation of this part 46 is completed, a re-initializing part 47 initializes (formats) the medium of the disk device where the error occurs. After initialization of the re-initializing part 47 is completed, a medium check part 48 checks the medium of the disk device where the error occurs. A second data restoration part 49 restores data of the auxiliary disk device 59 into an error disk device when it is discriminated by the medium check part 48 that the medium is normal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error recovery device for a disk array device which restores data to a spare disk device when a failure occurs in the disk device for data storage or parity storage, and more particularly to a spare disk device. The present invention relates to an error recovery device of a disk array device that attempts recovery operation of an error disk device after switching.

[0002]

2. Description of the Related Art In recent computer systems that are becoming faster and higher in performance, the performance of a central processing unit has been remarkably improved due to the progress of semiconductor technology. The same high performance is required for the. In response to this request, there is a limit to the speedup of the magnetic disk device that involves mechanical operation, so a disk array is configured with multiple magnetic disk devices and connected in parallel to the disk array control device, and multiple magnetic disk devices are 2. Description of the Related Art There is provided a disk array device that performs parallel access to perform read and write operations.

In such a disk array device, a spare disk device is provided for the operating disk device, and the spare disk device is switched to the spare disk device when the operation disk device fails. FIG. 6 shows a conventional disk array device. The disk array controller 2 is composed of a host device interface controller 3 connected to the host device 1 and a device controller 4 connected to the plurality of magnetic disk devices 50 to 59 of the disk array 5. The disk array 5 has data storage disk devices 50 to 57 and a redundant information storage disk device (hereinafter referred to as “redundant disk device”) 57, and further has a spare disk device 59.

The disk array control device 2 is a host device 1.
In response to a data transfer request from the magnetic disk devices 50 to 58, the magnetic disk devices 50 to 58 are accessed in parallel via the device control unit 4 to simultaneously perform the read process or the write process. That is,
When data is written to the plurality of data storage disk devices 50 to 57, parity data or the like is generated and written to the redundant disk device 58. Parity is the data from the data of other normal disk units and the parity data of redundant disk units even if some failure occurs in one magnetic disk unit in multiple data storage disks when reading data. It is possible to restore.

Further, when a failure occurs in one of the disk devices 50 to 58 in succession, the error disk device in which the error has occurred is separated from the logical disk allocation according to the instruction from the device control unit 4. All the data of the error disk device is restored to the newly allocated spare disk device.

The data restoration process for the spare disk device can be instructed by an operator, but normally, the disk array controller 2 monitors the error occurrence status,
Data restoration is started automatically when an error exceeds a certain value.

[0007]

By the way, most of the failures that occur in the disk device are due to deterioration and defects of the parts that make up the disk device, but due to the characteristics of the materials that make up the parts and the track deviation. Uncorrectable data checks may occur. Generally, these data check failures can be recovered by the initialization process (format process) of the medium.

However, in the conventional apparatus, it is premised that the failed disk device is replaced with a new one, and even if the failure is recoverable by the initialization processing of the medium, the failed disk must always be removed from the system case. Remove the device,
Since it was necessary to set it in another test device, etc. to initialize it, it required replacement of the disk device and manual work for initialization, so it took time to restore the disk array device. there were.

According to the present invention, the recovery process by the media initialization of the failed disk device is automatically performed without the need for manual or replacement of the disk device, and the recovery work for the failure occurrence is made efficient and processed in a short time. It is an object of the present invention to provide an error recovery device for a disk array device that is made possible.

[0010]

FIG. 1 is a diagram illustrating the principle of the present invention. First, the present invention relates to a data storage disk device 5
0 to 57 and a disk array 5 having a redundant information storage disk device 58 are connected to each other, and a disk array control device 2 for parallelly accessing a plurality of magnetic disk devices 50 to 58 in response to an access from the host device 1 is provided. Further, the disk array 5 is intended for a disk array device having at least one spare disk device 59.

In the present invention as an error recovery device for such a disk array device, the disk array control device 2 includes a first data restoration unit 46 and a re-initialization unit 4.
7. A medium inspection unit 48 and a second data restoration unit 49 are provided. The first data restoration unit 46, when the number of error occurrences of any one of the data storage disk devices 50 to 57 and the redundant information storage disk device 58 of the disk array 5 exceeds a specified value, the data of the error occurrence disk device. Are restored to the spare disk device 59. Re-initialization part 47
Initializes (formats) the medium of the error disk device after the restoration operation by the first data restoration unit 46 is completed. The medium inspection unit 48 inspects the medium of the error disk device after the initialization by the re-initialization unit 47 is completed. The second data restoration unit 49 restores the data of the spare disk device 59 to the error disk device when the medium inspection unit 48 determines that the medium is normal.

Further, the disk array control device 2 is provided with an upper report section 38, start and end of data restoration by the first data restoration section 46, start of re-initialization by the re-initialization section 47, and medium normality judgment by the medium inspection section 48. Of re-initialization based on the second data restoration unit 4
Each of the start and end of the data restoration by the 9
To notify.

Further, the upper report unit 38 reports the completion of the data restoration by the first data restoration unit 46, the re-initialization by the re-initialization unit 47, and the data restoration by the second data restoration unit 49 to the higher-level device 1. Monitors the elapsed time from the
Alternatively, if there is no instruction from the operator, the process is forced to shift to the next process.

Further, the disk array control device 2 is provided with a logging processing unit 39 for reporting data restoration by the first data restoration unit 46, re-initialization by the re-initialization unit 47, and data restoration by the second data restoration unit 49. The content is stored and held in the nonvolatile storage unit.

[0015]

According to the error recovery device of the disk array device of the present invention, the following effects can be obtained. When an error occurs frequently in either the data of the disk array or the disk device for redundant recording and the number of error occurrences exceeds a specified value, the data restoration operation to the spare disk device is automatically started. At this time, the start of data restoration is notified to the host device. When the data restoration operation to the spare disk device is completed, the fact is reported to the upper device and the instruction for the completion report is waited by the time monitoring.

When there is an instruction from the host device or the operator or when the monitoring time overflows, the error disk device is reinitialized. When the re-initialization is completed, an inspection process (diagnosis process) for inspecting the medium that has been initialized next is performed. If the medium diagnosis is normal, the completion of re-initialization is reported to the host device at this point, and the instruction for the completion report is waited by time monitoring.

When there is an instruction from the host device or an operator, or when the monitoring time overflows, data is restored from the spare disk device to the disk device in which the error has been recovered by re-initialization, and the host device is notified of the completion of data restoration. To do. As a result, the failure of the disk device that can be recovered by the initialization of the medium can be recovered to the normal disk device without replacing the disk device or requiring manpower.

Even if the disk array control device is disconnected from the host device, the logging processing unit stores the recovery status and the result of the failure occurrence in the non-volatile storage unit of the disk array control device, and provides it to the host device as logging information. be able to.

[0019]

FIG. 2 is a block diagram showing an embodiment of the present invention. 2, the disk array device of the present invention comprises a disk array control device 2 connected to a host computer 1 as a host device and a disk array 5 in which a plurality of disk devices 50 to 59 are connected in parallel as logical devices. It In this embodiment, the disk array 5 includes eight storage disk devices 50 to 57 for storing data, one redundant disk device 58 for storing parity information, and one spare disk device. It is composed of 59.

The disk array device 2 comprises an upper interface control section 3 connected to the host computer 1 and a device control section 4 connected to the disk array 5. The upper interface control unit 3 is provided with an interface control unit 31, an MPU 32, a data transfer control unit 33, a flag register 35, a counter 34, and a non-volatile storage unit 36.

The MPU 32 performs various processes in response to a data transfer request from the host computer 1 by the microprogram 37, and among its processing functions, various states associated with the state of the disk array 5 by the device control unit 4, particularly error recovery processing. Function of the upper report unit 38 for reporting the status and result of the above to the host computer 1, and the logging processing unit 39 for storing and retaining the error recovery status and result reported by the upper report unit 38 in the nonvolatile storage unit 36 as logging information. Is provided. Further, an operator control unit 6 is connected to the MPU 32 so that the operator can send information necessary for various maintenance such as error recovery from the operator control unit 6 to the M unit 32.
The PU 32 can be instructed.

The device control section 4 is provided with a disk array control section 41, an MPU 42, a data transfer control section 43, and a data check counter 44. The MPU 42 executes the microprogram 45 to perform a read operation or a write operation with respect to the disk array 5 in response to a data transfer request from the host computer 1 by the MPU 32 of the host interface control unit 3, and further an error recovery processing operation of the present invention. To do.

For this error recovery, the microprogram 45 is provided with the respective functions of a first data restoration section 46, a reinitialization section 47, a medium inspection section 48 and a second data restoration section 49. Data check counter 44
Are disk devices 50 to 59 provided in the disk array 5.
Each has a counter area, and the read data obtained by the read operation at the time of accessing the disk array in response to the data transfer request from the host computer 1
When an uncorrectable error is detected by C, it is determined that a failure has occurred, and the value of the data check counter 44 corresponding to the disk device in which the error has occurred is incremented by one.

The first data restoring unit 46 monitors the count value of the data check counter 44, and when the number of error occurrences reaches a preset specified value, the disk device whose error count has reached the specified value is set as an error disk. It is determined that the device is a device and is designated as the target of the error recovery process, and the data restoration process for restoring the data of the error disk device to the spare disk device 59 is executed.

Data restoration to the spare disk device 59 can be generated by using each data of the normal storage disk device except the error disk device and the redundant disk device 58. The re-initialization unit 47 gives an instruction from the host computer 1 or the operator control unit 6 when the data restoration of the error disk device to the spare disk device 59 is normally completed by the first data restoration unit 46, or when there is no instruction. The counter 34 provided in the host interface control unit 3 is activated when the time is monitored and overflows, and reinitialization of the medium of the error disk device, that is, formatting as an initialization process is executed.

The medium inspecting unit 48 includes a re-initializing unit 4
Error due to 7) Start up at the point when initialization of the medium of the disk device is completed, write predetermined dummy data on the entire data surface of the initialized medium, and then read the entire surface. Perform the medium inspection. When the inspection by the medium inspection unit 48 is normally completed, the re-initialization is completed. Completion of the re-initialization is reported to the host computer 1 and the operator control unit 9.

The second data restoring section 49 receives an instruction from the host computer 1 or the operator control section 6 or the upper interface control section 3 after the completion of the re-initialization.
When the time monitoring by the counter 34 provided in 1 is overflowed, the data is stored in the spare disk device 59 for the disk device in which an error has occurred, which is activated when it is reinitialized and can operate normally. In this case, since the spare disk device 59 is operating normally, the data of the spare disk device 59 is copied to the disk device for which error recovery has been completed.

Furthermore, the MP of the upper interface control unit 3
The upper report unit 38 provided as a function of U32 starts and ends the error recovery process by the first data restoration unit 46, the re-initialization unit 47, the medium inspection unit 48, and the second data restoration unit 49 by the MPU 42 of the device control unit 4. And the result is reported to the host computer 1. Regarding the re-initialization, the start of the re-initialization is reported by the re-initialization unit 47, and the end of the re-initialization is normally terminated by the medium inspection unit 48, and the completion of the re-initialization is reported.

The upper report section 38 is provided for the host computer 1
In addition, the status and result of the error recovery processing can be reported to the operator control unit 6 as necessary. For example, when a maintenance person is attached to the disk array control device 2, the operator is notified of the situation and the operation panel or the like is used to display a status or a result with a predetermined code number to give an instruction to the operator for error recovery. Can wait

Further, when the upper reporting unit 38 reports the start of various operations for error recovery to the host computer 1, it activates the counter 34 and monitors the time, and the count value of the counter 34 remains constant for a certain period of time. If an overflow occurs later, the MPU 42 is instructed to shift to the next error recovery process without waiting for an instruction from the host computer 1 or the operator control unit 6.

The reporting process to the host computer 1 by the upper report unit 38 is performed according to the state of the flag register 35. When the flag register 35 is set to 1, the upper report unit 38 reports to the host computer 1 by interrupt processing. On the other hand, when the flag register 35 is reset to 0, the response to the access from the host computer 1 is reported to the host device.

That is, when the disk array controller 2 is disconnected from the host computer 1, the flag register 35 is set to 1. In this state, the interrupt is reported to the host computer 1.
On the other hand, when the host computer 1 and the disk array control device 2 are connected and data is being transferred, for example, the status information at the end of the transfer is included in the report to the higher-level device.

FIG. 3 shows the disk array controller 2 of FIG.
2 is a schematic diagram of a data transfer process by. First, step S1
Then, the MPU 32 of the host interface control unit 3 checks whether or not there is an input / output request from the host computer 1 due to data transfer. When there is an input / output request, the process proceeds to step S2, where a read command or a write command is issued to the MPU 42 of the device control unit 4, and the storage disk devices 50 to 57 of the disk array 5 are issued via the disk array control unit 41. Further, the read operation or the write operation of step S2 is performed by the parallel access of the redundant disk device 58.

For example, in response to a write data transfer request from the host computer 1, the data is stored via the channel interface controller 31, the data transfer controller 33, the data transfer controller 43, and the disk array controller 41. Data is written to the disk devices 50 to 57 and parity data is written to the redundant disk device 58.

Further, in response to a read data transfer request from the host computer 1, data is read from the storage disk devices 50 to 57 of the disk array 5, and the disk array controller 41, the data transfer controller 43, and the data are transferred. Transfer control unit 33, channel interface control unit 31
The requested data is transferred to the host computer via.

Next, in step S3, it is checked whether or not there is a disk in which an uncorrectable error has occurred in the disk device in operation of the disk array 5. If there is a disk device that has generated an uncorrectable error, the process proceeds to step S4, and the MPU 42 increments the error occurrence count of the corresponding counter area of the data check counter 44 by one.

Next, in step S5, it is checked whether or not there is a disk device having the number of error occurrences exceeding the predetermined value in the value of the data check counter 44. If there is a disk device with the number of error occurrences exceeding the specified value, the disk device is determined to be an error disk and the process proceeds to the error processing of step S6. 4 and FIG.
4 is a detail of the error processing according to the present invention in step S6 of FIG. Regarding this error processing, the disk array 5 of FIG.
An example will be described in which the number of error occurrences of the storage disk device 50 provided in FIG.

In the MPU 42, when the value of the data check counter 44 of the storage disk device 50 reaches a specified value, it is determined that the disk is an error disk and a failure notification report is sent to the MPU 32. MPU32 that received this failure notification report
Instructs the MPU 42 to start a data restoration process for restoring the data of the error disk device 50 to the spare disk device 59, as shown in step S1 of FIG.

At the same time, the MPU 32 reports to the host computer 1 that the data restoration process has been started by the function of the upper report unit 38, as in step S2. At this time, the MPU 32 checks the state of the flag register 35, and if the flag register 35 is set to 1, it reports the start of the data restoration process to the host computer 1 by an interrupt, while the flag is reset to 0. If this is the case, the start of the data restoration process is reported by including it in the status information that accompanies the current access from the host computer 1.

The MPU 42, which has received the data restoration start instruction from the MPU 32, restores the data of the error disk device 50 to the spare disk device 59 via the disk array control unit 41 by the function of the first data restoring unit 46. Start the restoration process of. This data restoration process can be generated based on each data of the normal storage disk devices 51 to 57 excluding the error disk device 50 and the parity data of the redundant disk device 58.

When all the data in the error disk device 50 for the spare disk device 59 has been restored and normal termination is determined in step S3, the process proceeds to step S4 and the MPU
42 reports the completion of data restoration to the MPU 32. In response to this, the MPU 32 receives the flag register 35 at that time.
The data restoration completion report to the host computer 1 is issued while referring to the state of.

After completing the data completion report to the host computer 1, the MPU 32 waits for the confirmation response from the host computer 1 in step S5, and judges that the report is completed only when the confirmation response is obtained, and the next step S5.
Proceed to 6. While waiting for the confirmation response from the host computer 1, the logging processing unit 39 is activated in step S6, and an internal logging process for recording the state of data restoration completion for the spare disk device 59 in the non-volatile storage unit 36 is performed.

When the confirmation response is normally obtained from the host computer 1 and the report is completed in step S5, the process proceeds to step S6, in which the MPU 32 activates the counter 34 and starts the time monitoring. The counter 34 overflows after a lapse of a predetermined time, and indicates that the monitoring time has ended. If there is a re-initialization instruction from the host computer 1 or the operator control unit 6 within the monitoring time when the counter 34 overflows, the process proceeds to the next step S7. Even if there is no instruction for re-initialization, the MPU will be released when the counter 34 overflows.
42 will be instructed to re-initialize.

The MPU 42 receives a reinitialization instruction based on an instruction from the host computer 1 or the operator control unit 6 by the MPU 32, or a reinitialization instruction based on the overflow of the counter 34 when there is no instruction. Fifty
To re-initialize the medium. In response to this instruction, the error disk device 50 starts the initialization operation for re-forming the formatting of the medium again via the disk array control unit 41 as in the factory shipment.

When it is determined in step S8 that the reinitialization of the error disk device 50 is normally completed, the MPU4
In step S9, the dummy disk 2 is written to the entire data area of the re-initialized medium, and then the entire surface of the error disk device 50 is read.
Instruct to start the medium inspection process for checking the read result.

Subsequently, in step S10, the MPU 42 indicates that the medium inspection process in the error disk device 50 is normally completed.
When the determination is made in step 3, the MPU 42 reports the completion of the re-initialization process to the MPU 32. In response to this, MPU32
Responds to the host computer 1 according to the state of the flag register 35 at that time to report the completion of the re-initialization process as in step S11.

In response to the completion report of the re-initialization process,
In the next step S12, it is monitored whether or not there is a confirmation response from the host computer 1, and in the meantime, in step S12.
In step 22, the completion of the re-initialization process is confirmed by the nonvolatile storage unit 3
6 is stored and held as internal logging processing. When the completion of the report is determined in step S12 in response to the confirmation response from the host computer 1, the MPU 32 is determined in step S13.
Resets the counter 34 and restarts it, and starts time monitoring for receiving an instruction from the host computer 1 or the operator control unit 6. If there is an instruction before the counter 34 overflows, step S14 in FIG.
Proceed to. Even if there is no instruction, if the counter 34 overflows after a certain time in step S23, the process proceeds to step S14 in FIG.

In step S14 of FIG. 5, even if there is no instruction from the host computer 1 or the operator control section 6 or there is no such instruction, the disk device in which the error that the re-initialization is normally completed is caused based on the overflow of the counter 34. Spare disk device 5 for 50
The data restoration instruction from the MPU 9 is given to the MPU 42, and the data restoration process is started.

Then, when the MPU 42 determines in step S15 that the error recovery of the data of the spare disk device 59 to the disk device 50 is normally completed, the MPU 32 is notified of the normal completion of the data recovery process. MPU32
Depending on the state of the flag register 35 at that time
The completion report of the recovery process is performed as in step S16.

Subsequently, in step S17, the confirmation response from the host computer 1 is waited, and in the meantime, in step S25, the disk device 50 in which the error has occurred is recovered in the non-volatile storage unit 36 and the data restoration is completed. Performs an internal logging process that records Host computer 1
A confirmation response to the completion report of the restoration process is sent in step S.
When the determination is made in 17, the recovery process associated with the occurrence of a series of errors is ended, and the process returns to the main routine of FIG.

On the other hand, in step S3, when the data restoration of the error disk device 50 to the spare disk device 59 cannot be normally completed, there is a failure in the spare disk device 59, and the process proceeds to the error processing of step S18. In this case, the spare disk device 59 is replaced in addition to the error disk device 50, and the necessary data restoration processing is performed.

If the re-initialization does not end normally in step S8 or the medium inspection process does not end normally in step S10, the error disk device 50 cannot be used even if re-initialization is performed in step S21. The error processing is performed by replacing the error disk device 50. Further, in step S15, if the data restoration from the spare disk device 59 to the disk device in which the error occurred after the completion of the re-initialization cannot be normally completed, the disk device 50 cannot be recovered by the re-initialization in step S24. It is determined that another failure has occurred, and error processing for replacing the disk device 50 is performed.

Although the above-mentioned embodiment has exemplified the disk array using the magnetic disk device, it can be applied to an array device using an appropriate physical device such as an optical disk device and a semiconductor memory device. Moreover, the number of storage disk devices provided in the disk array 5 can be appropriately determined as necessary. Although the disk array 5 of the embodiment has a one-rank configuration as an example, it may have a multi-rank configuration in which a parallel configuration is provided in multiple stages.

Further, in the above embodiment, the host computer 1 starts the data restoration from the error disk unit to the spare disk unit and the data restoration from the error disk unit to the spare disk unit is completed. The error disk unit is reinitialized. Completion of processing Although the completion of data restoration from the spare disk unit is reported to the error disk unit, it is sufficient if at least the first data restoration start report and the completion of the last recovery process are reported, and the report between them is necessary. It can be appropriately determined depending on the situation.

In particular, in the present invention, a report is sent to the host device and an instruction is awaited. However, even if there is no instruction, it is possible to automatically shift to the next error recovery processing by the time monitoring due to the overflow of the counter. It is not necessary to report the status to the host device. However, in order to prevent the state of the disk array control device 2 side from becoming invisible to the host device,
At least the nonvolatile storage unit 36 needs to store the error recovery logging information.

[0056]

As described above, according to the present invention, when the data restoration of the disk device in the disk array is started due to the occurrence of the uncorrectable data check, the internal processing of the disk array controller is started. This automatically performs recovery work including re-initialization to return the failed disk device to the usable state as much as possible without the need for work instructions from the host device or operator. The restoration can be performed without delay, and further, it is possible to prevent an operation error by human.

For the disk in which the error has occurred, a media inspection is automatically performed by the re-initialization and the full read operation after the completion of the re-initialization, and it is assumed that the error has been recovered by the normal end, and the data of the spare disk device is restored. As a result, the original operating state is automatically restored, and error recovery processing can be efficiently performed against the occurrence of a data check that is recovered by initializing the medium.

[Brief description of drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is a block diagram showing an embodiment of the present invention.

FIG. 3 is a schematic flowchart of an access process of the present invention.

FIG. 4 is a detailed flowchart of error processing in FIG.

FIG. 5 is a detailed flowchart of error handling in FIG. 3 (continued).

FIG. 6 is a block diagram of a conventional device.

[Explanation of symbols]

 1: Host device (host computer) 2: Disk array control device 3: Host interface control unit 4: Device control unit 5: Disk array 6: Operator control unit 31: Interface control unit 32, 42: MPU 33, 43: Data transfer Control unit 34: Counter 35: Flag register 36: Non-volatile storage unit 37, 45: Micro program 38: High-order report unit 39: Logging processing unit 41: Disk array control unit 44: Data check counter 46: First data restoration unit 47: Reinitialization unit 48: Medium inspection unit 49: Second data restoration unit

Claims (4)

[Claims] 1. A disk array controller for connecting a disk array having a plurality of disk devices for storing data and storing redundant information, and for parallelly accessing the plurality of magnetic disk devices in response to an access from a host device. The disk array further comprises at least one spare disk device, wherein the disk array control device includes a plurality of disk devices for storing data of the disk array and storing redundant information. When any one of the error occurrence counts exceeds the specified value, the first data restoration unit restores the data of the error occurrence disk device to the spare disk device, and after the restoration operation by the data restoration unit is completed, A re-initialization unit that initializes the medium of the error disk device, and a re-initialization unit. A medium inspecting unit for inspecting the medium of the error disk device after the initialization is completed, and for restoring the data in the spare disk device to the error disk device when the medium inspecting unit determines that the medium is normal. An error recovery device for a disk array device, comprising: a data recovery unit. 2. The error recovery device for a disk array device according to claim 1, further comprising the start and end of data restoration by the first data restoration unit, the start of re-initialization by the re-initialization unit, and the medium inspection. End of re-initialization based on medium normality judgment by the section, the second
Each of the start and end of data restoration by the data restoration unit,
An error recovery device for a disk array device, which is provided with an upper report unit for notifying the upper device. 3. The error recovery device for a disk array device according to claim 2, wherein the higher-order reporting unit is configured to restore data by the first data restoration unit, re-initialize by the re-initialization unit, and the second data. For each data restoration by the restoration unit, monitor the elapsed time from the completion report to the upper device, and if there is no instruction from the upper device or operator even after a certain period of time, forcibly move to the next process An error recovery device for a disk array device, characterized by: 4. The error recovery device for a disk array device according to claim 1, further comprising: data restoration by the first data restoration unit, re-initialization by the re-initialization unit, and data by the second data restoration unit. An error recovery device for a disk array device, which is provided with a logging processing unit for storing and holding each report content of restoration in a non-volatile storage unit.