JP2880000B2 - Array type disk drive system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the field of disk storage for electronic computers, and more particularly to an array type disk drive system for exchanging data with a host computer.

[0002]

2. Description of the Related Art Generally, in a disk drive system for exchanging data with a host computer, in order for a user to use the system continuously for a long period of time, the disk drive mechanism is hardly broken down with reliability. Even if a part of the disk fails and becomes inaccessible, the ability to reproduce the lost data in the disk in some way is required.

Generally, in order to restore the contents of a lost disk so that it can be used again,
There is a method in which the system is temporarily stopped, and all efforts are made to restore / copy the contents. However, if the system is shut down, users of the system will incur disadvantages during that time because the system is unusable. Therefore, in order to avoid the above disadvantage, it is necessary to restore the contents without stopping the system, that is, online.

Conventionally, a RAID (Redundant Array) proposed by the University of California, Berkeley (UCB) has been proposed to meet such demands.
This RAID aims at high reliability by arranging a large number of small-capacity and inexpensive disks and providing a large-capacity and high IOPS and further having redundancy information.

[0005] The UCB paper "A Case fo
r Redundant Arrays of Ine
xpensive Disks (RAID) ", Da
vid A. Patternson, Garth G
ibson, and Randy H .; Katz, R
report No. UCB / CSD 87/391, D
In 1987, RAID1 is changed to RAID1.
５5 levels. As a specific embodiment, an array type disk drive mechanism system as shown in FIG. 4 has been proposed.

[0006] The contents shown in FIG.
No. 4 discloses a preferred embodiment (with some modifications) of the array-type disk drive mechanism system. In the figure, reference numeral 21 denotes an array of individual disks 20. The array 21 is arranged to have eleven vertical channels 19. Each of these channels 19 has six disks 20.
The disk 20 is of a magnetic type or an optical type.
Access to one disk 20 is executed by the channel controller 18.

[0007] On the data bus 16 is an error correction and control (ECC) engine 17 which provides redundancy for the data stored on the disk 20 with the software of the arrayed disk drive system 22. The memory buffer 15 is arranged on the data bus 16,
When data is written from the host computer 13 at the time of a write request, or when data is read from the disk 20 in response to a read request from the host computer 13, the data is stored in the memory buffer 15. Also, a microprocessor (in the above example, R
An ISC processor 24 controls the entire array disk controller.

[0008] The redundancy information is necessary for reproducing lost data. This redundancy information is obtained by taking the exclusive OR of each block of the same redundancy group 23 for each bit, thereby performing error correction and control (ECC).
Generated by the engine 17. This is called parity. The redundancy group 23 is one unit for reproducing data blocks. In the case of FIG.
Since one redundancy group 23 is formed, the number of data blocks is 10, and the parity generated by taking the exclusive OR of them is present in the remaining one disk 20.

Therefore, if data on a certain disk 20 is lost due to some kind of failure, reproducing the lost data is a simple process if the data and parity of the other disks 20 are known. That is,
The exclusive OR of all surviving data and parity may be calculated for each bit.

Further, as a reference technical document related to the present invention, there is a "failure recovery type storage subsystem" disclosed in Japanese Patent Application Laid-Open No. 2-135555.

[0011]

In the conventional array type disk drive system, the simple reproduction principle of lost data as described above is described. However, any specific data correction / restoration is performed. There was a problem that the method was not described.

According to the present invention, when one of the disks in the redundancy group fails, the disk is modified / restored online to improve the operability of the system, reduce the amount of memory, and develop / maintain (maintain). It is an object of the present invention to obtain an array type disk drive mechanism system that is easy to use.

[0013]

An array-type disk drive system according to the present invention has a spare disk drive in addition to a plurality of disk drives for storing data and parity as array elements. In an array-type disk drive system operating as one logical disk in response to an access request from a host computer, boundary information for dividing a storage area of the spare disk drive into a restored area and an unrestored area is stored. Boundary information storage means for, and said spare disk drive mechanism,
In order to operate as a substitute for a failed disk drive among the plurality of disk drives, data that should be stored in the failed disk drive is stored in a predetermined order in the spare disk drive. A restoring means for storing the boundary information indicating the end position of the storage area in which the data storage is completed in the boundary information storage means, and storing the upper level data during the operation of the restoring means. When there is an access request from the computer, it is determined whether or not the access request is related to the data in the restored area based on the boundary information stored in the boundary information storage means. If so, the spare disk drive responds to the access request as having been restored, otherwise, Serial spare disk drive mechanism in which and a response means responsive to said access request as if not yet recovered.

[0014]

According to the present invention, the array disk controller detects a disk abnormality, switches the disk in which the abnormality is detected to a spare disk, and uses a new algorithm to correct / correct the data of the failed disk.
Perform recovery online to a spare disk.

[0015]

FIG. 4 is an overall configuration diagram of one embodiment of a disk drive mechanism system according to the present invention. this is,
It has the same configuration as a conventionally proposed disk drive mechanism system. In the figure, if one of a plurality of arrays 21 or one of the disks 20 in an array 21 becomes inaccessible for some reason, the disk is automatically switched to a spare disk and the data is stored in the following procedure. Modify / Restore.

For example, a case where a spare disk is used in RAID5 will be described. As shown in FIG. 2, the present algorithm restores the contents of the failed disk on the spare disk in the order of addresses, that is, in order from the top in the figure. During this process, when an access command (read / write command) to the disk being modified / restored is received, processing is performed according to which part of the disk being modified / restored as described below. And That is,
Based on the "boundary" of restored / unrestored, the required range of the access instruction is classified according to the position.

Here, for convenience of explanation, the array 1
One is assumed to be one large capacity disk, and the number of arrays is assumed to be five. In the following first embodiment, "boundary"
It is characterized by using the concept of

First, the case where a read / write command comes to the restored portion 6 of the disk being modified / restored shown in FIG. 2 will be described. In this case, the operation is the same as the normal operation.
First, for the leads, for example, e4 to e4 in FIG.
When an instruction to read data in the range of e7 comes,
The parity of e6 is skipped, and the contents of e4, e5, and e7 are temporarily written in the memory buffer 15 shown in FIG. This will be referred to as “normal read operation”.

The following is a description of the light. However, new data and parity to be written have n
(New). Write request e
4 to 7 (excluding the parity e6), and furthermore, the parity of the same redundant column, that is,
b4 is read for e4, a5 is read for e5, and d7 is read for e7, and written to the memory buffer 4 shown in FIG.

Next, the same redundant columns, that is, e
4 and b4 and ne4, e5 and a5 and ne5, and e7 and d7 and ne7 perform an exclusive OR operation on each bit (this operation is executed by the error correction and control (ECC) engine 17), and the new parity nb4 , Na5, nd7
Is generated and written into the memory buffer 15 shown in FIG. Finally, new data, parity ne4, ne5,
Ne7, nb4, na5, and nd7 are written to the respective disks. This will be referred to as “normal write operation”.

Secondly, a case where a read / write command comes to the unrestored portion 7 of the disk being modified / restored shown in FIG. 2 will be described. In this case, it is necessary to first restore the data for which the read / write request has been made. For example, when a read / write command arrives at e9 to e13 in FIG.
Data and parity of the same redundant column as the respective data e9, e10, e12 and e13, that is, a9 and b9, c9 and d9 for e9, and a10 and b for e10
A12 and b12 for 10, c10 and d10, e12
, C12 and d12, and a13, b13 and c for e13
13 and d13 are read and written to the memory buffer 15.

Next, an exclusive OR is calculated for each bit in each redundant column and written to the memory buffer 15. The calculation of the exclusive OR is performed by an error correction and control (ECC) engine 17. As a result, the data in the requested range, that is, the data in the portions e9 to e13 are reproduced. Next, in the case of a read, the reproduced data is transferred to the host computer 13. This operation of reproduction and transfer will be referred to as “corrected read operation”.

In the case of a write, the old data (e9, e9,
e10, e12, e13) and new data (ne9, n)
e10, ne12, ne13) and old parity (b
9, a10, d12, c13), an exclusive OR is calculated for each bit, and a new parity (nb9, na10, n) is calculated.
d12, nc13) is generated and written to the memory buffer 15. Thereafter, only the new parity is written to each disk. Playback and this operation are collectively referred to as a "corrected write operation". Here, the reason why new data is not written is that the concept of “boundary” is broken.

Third, a case where a read / write instruction comes to a location that covers both a restored portion and an unrestored portion will be described. However, the lead / non-restored portion is read in contact with the boundary line 8.
When a write instruction comes, that is, e8 to e12 in FIG.
In this case, the case where a read / write instruction comes in is also included.

In this embodiment, each location is processed differently. That is, for both read / write,
The restored portion is processed by the "normal read / write operation", and the unrestored portion is processed by the "corrected read / write operation".

In the first embodiment, the restored portion is processed by the "normal read / write operation", and the unrestored portion is processed by the "corrected read / write operation". In the second embodiment, the above processing is performed by another method.

For example, a case where a read / write instruction is received at e5 to e9 shown in FIG. 3 will be described. In the second embodiment, without executing the read / write instruction for the time being,
Focus on restoring the unrestored portion of the requested range.
In other words, in this example, the e8 and e9 are reproduced and then written to the disc. This allows
The restored / unrestored boundary line advances to the position of e9.

Thereafter, the read / write instruction is restarted when writing is completed up to e9. That is, here, a read / write command is sent to a location that has already been restored, so that it is processed by a "normal read / write operation".

The third embodiment is performed by a method different from the method shown in the first embodiment. For example, e5 in FIG.
A description will be given of the case where the read / write instruction of ~ e9 comes. When the read instruction comes, it is the same as the method of the second embodiment described above. That is, the "normal read operation" is performed after restoring the part up to which the request has been received.

Next, when a write instruction is received,
In this case, e5 and e7 are read out together with the parity in the same redundant column and stored in the memory buffer 15,
e8 and e9 are reproduced and then stored in the memory buffer 15. In addition, new data ne5
ne7, ne8, ne9, old data e5, e7, e8,
e9, new parities na5, nd7, nc8, and nb9 are generated from the old parities a5, d7, c8, and d9 by exclusive ORing the same redundant columns for each bit. After that, the new data and the new parity are respectively written to the disk. This will advance the boundary line.

In each of the above embodiments, the error correction and control (ECC) engine is used to generate the redundancy information. However, software may be used without using such a circuit. Further, even if the hardware configuration is not the one used in the description, if it is an array type disk drive system having redundancy information and a spare disk,
The present invention can be applied.

FIG. 1 is a conceptual diagram showing the outline of the operation of an array type disk drive mechanism system when the method of the present invention is used. The dashed arrow for the spare disk drive mechanism 3 is due to consideration of writing to the unrestored portion.

FIG. 5 is a flowchart showing the operation of the first embodiment. First, it is determined whether or not the restoration has been completed (S20). As a result, if it is determined that the processing has been completed, the processing ends. Conversely, if it is determined that the restoration has not been completed, it is next determined whether or not there is an I / O request (S21). As a result, when it is determined that there is no I / O request, the restoration operation is executed in a certain unit (S2).
2) Return to step 20 above.

On the other hand, if it is determined that there is an I / O request, it is determined whether the request is for the disk being restored (S23). As a result, when it is determined that the data is not for the disk being restored, read / write determination is performed (S24), and then a normal read / write operation is performed (S25), and the process returns to step 20. Conversely, if it is determined that the data is for the disk being restored, the process proceeds to step 26.

That is, it is determined whether the data is read or write (S
26) If it is determined to be a read, it is next determined to which part to read (S27). If it is determined that the restored unit is one, the normal read operation is performed (S28), and the process returns to step S20. If it is determined that the part is the second unrestored part, a modified read operation is performed (S29), and the process returns to step S20. Further, when it is determined that both are the three, the normal read operation is performed on the restored part (S30), and then the corrected read operation is performed on the unrestored part (S31), and the process returns to the step S20.

If it is determined in step 26 that the data is a light, the flow shifts to the flowchart shown in FIG. 6 to determine to which part the data is to be written (S32). Where 1
If it is determined that the part is the restored part, the normal write operation is performed (S33), and the process returns to step S20. If it is determined that the part is the second unrestored part, a correction write operation is executed (S34), and the process returns to step S20. In addition, 3
If it is determined that both are the same, a normal write operation is performed on the restored portion (S35), and then a modified write operation is performed on the unrestored portion (S36), and the process returns to step 20.

FIG. 7 is a flowchart showing the operation of the second embodiment. First, it is determined whether or not the restoration has been completed (S40). As a result, if it is determined that the processing has been completed, the processing ends. Conversely, if it is determined that the restoration has not been completed, it is next determined whether or not there is an I / O request (S41). As a result, when it is determined that there is no I / O request, the restoration operation is executed in a certain unit (S4).
2) Return to step 40 above.

On the other hand, when it is determined that there is an I / O request, it is determined whether the request is for the disk being restored (S43). As a result, if it is determined that the data is not for the disk being restored, read / write determination is performed (S44), and then a normal read / write operation is performed (S45), and the process returns to step 40. Conversely, if it is determined that the data is for the disk being restored, the process proceeds to step 46.

That is, it is determined whether the data is read or write (S
46) If it is determined that it is a read, it is next determined to which part the read is to be performed (S47). If it is determined that the restored unit is one, the normal read operation is performed (S48), and the process returns to step S40. If it is determined that the part is the second unrestored part, a modified read operation is performed (S49), and the process returns to step S40. Further, when it is determined that both of the three are satisfied, the normal read operation is executed after restoring the unrestored portion (S50), and the process returns to step 40.

If it is determined in step 46 that the data is a light, the flow shifts to the flowchart shown in FIG. 8, and it is determined to which part the data is to be written (S52). Where 1
If it is determined that the part has been restored, a normal write operation is executed (S53), and the process returns to step S40. If it is determined that the part is the unrestored part, a modified write operation is executed (S54), and the process returns to step S40. In addition, 3
If it is determined that both are the same, after restoring the unrestored portion (S55), a normal write operation is executed (S56), and the process returns to step 40.

FIG. 9 is a flowchart showing the operation of the third embodiment. First, it is determined whether or not the restoration has been completed (S60). As a result, if it is determined that the processing has been completed, the processing ends. Conversely, if it is determined that the restoration has not been completed, it is next determined whether or not there is an I / O request (S61). As a result, when it is determined that there is no I / O request, the restoration operation is executed in a certain unit (S6).
2) Return to step 60 above.

On the other hand, if it is determined that there is an I / O request, it is determined whether the request is for the disk being restored (S63). As a result, when it is determined that the data is not for the disk being restored, a read or write determination is performed (S64), and then a normal read / write operation is performed (S65), and the process returns to step 60. Conversely, if it is determined that the data is for the disk being restored, the process proceeds to step 66.

That is, it is determined whether the data is read or write (S
66) If it is determined that it is a read, it is next determined to which part it is to be read (S67). If it is determined that the restored unit is one, a normal read operation is performed (S68), and the process returns to step S60. If it is determined that the part is the second unrestored part, a modified read operation is executed (S69), and the process returns to step S60. Further, when it is determined that the number is both 3, after restoring the unrestored portion (S70), a normal read operation is executed (S71), and the process returns to step 60.

If it is determined in step 66 that the data is a light, the flow shifts to the flowchart shown in FIG.
It is determined to which part the light is to be written (S72). here,
If it is determined that the restored unit is No. 1, a normal write operation is performed (S73), and the process returns to step S60. When it is determined that the part is the second unrestored part, a correction write operation is executed (S74), and the process returns to step S60. further,
If it is determined that both are the three, the non-restored portion is reproduced (S75), and then the normal write operation is executed (S76).
The process returns to step 60.

On the other hand, a method that does not use the concept of “boundary” is also conceivable. That is, when a write command comes to the spare disk, the write is performed as it is. In this method, the restored portion is in a worm-eating state as shown in FIG. Therefore, although it is referred to as the worm-type correction / restoration, when a read instruction is received over a plurality of restored parts and unrestored parts, the control becomes complicated and the memory for managing the restored parts becomes enormous. Disadvantage. Compared to such a method, the method according to the present invention is simple and clear, and the development / maintenance (maintenance)
Is easy.

[0046]

As described above, according to the present invention, a failed disk can be corrected / restored online without stopping the system, so that the operability of the system can be improved.

In addition, since the concept of a boundary between a restored portion and an unrestored portion is provided and the correction / restoration is performed so as not to break it, the amount of memory used for managing the restored portion / unrestored portion is reduced. And development / maintenance (maintenance) is facilitated.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing an operation outline of an array type disk drive mechanism system when a method according to the present invention is used.

FIG. 2 is an explanatory diagram showing a disk drive mechanism that has been modified / restored to some extent by the method according to the present invention.

FIG. 3 is an explanatory diagram showing a disk drive mechanism which is being modified / restored or not being modified / restored by the method of the present invention.

FIG. 4 is an explanatory diagram showing an embodiment of a disk drive mechanism system according to the present invention and an embodiment of a conventional disk drive mechanism system.

FIG. 5 is a flowchart showing the operation of the first embodiment according to the present invention.

FIG. 6 is a flowchart showing the operation of the first embodiment according to the present invention.

FIG. 7 is a flowchart showing the operation of the second embodiment according to the present invention.

FIG. 8 is a flowchart showing the operation of the second embodiment according to the present invention.

FIG. 9 is a flowchart showing the operation of the third embodiment according to the present invention.

FIG. 10 is a flowchart showing the operation of the third embodiment according to the present invention.

FIG. 11 is an explanatory diagram showing the disk drive mechanism system being restored by the worm-eating type correction / restoration method.

[Explanation of symbols]

1 failed disk drive mechanism, 2 failed disk drive mechanism, 3 spare disk drive mechanism, 4 normal read / write operation, 5 read / write operation on spare disk drive mechanism, 6 restored part 7 Unrestored portion, 8 Boundary line between restored portion and unrestored portion, 9 Data block or parity block, 10 Disk drive mechanism being modified / restored, 11 Disk drive mechanism not being modified / restored, 13 Host computer, 15 Memory buffer, 17 error correction and control (ECC) engine, 18 channel controller, 19 channels, 20 disks, 21 arrays,
24 Microprocessor.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G06F 3/06 G06F 11/16 G06F 13/10

Claims (1)

(57) [Claims] A spare disk drive is provided in addition to a plurality of disk drives for storing data and parity as array elements, and operates as one logical disk in response to an access request from a host computer. An array-type disk drive mechanism system, comprising: boundary information storage means for storing boundary information for dividing a storage area of the spare disk drive mechanism into a restored area and an unrestored area; and In order to operate as a substitute for a failed disk drive among the plurality of disk drives, data that should be stored in the failed disk drive is stored in a predetermined order in the spare disk drive. And a boundary indicating the end position of the storage area where data storage is completed. Means for storing information in the boundary information storage means, during the operation of the recovery means, when an access request is received from the host computer, the access is performed based on the boundary information stored in the boundary information storage means. It is determined whether or not the request is related to the data in the restored area, and if so, the spare disk drive is regarded as being already restored and the A response unit that responds and otherwise responds to the access request assuming that the spare disk drive has not been restored.