JP2735071B2 - Defective block management method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a plurality of disk devices (magnetic disk devices, etc.) provided with a replacement area for a defective block (an area used as a replacement for the defective block when a defective block is detected) on a recording medium. The present invention relates to a defective block management method for managing defective blocks in a computer system including the same.

[0002]

2. Description of the Related Art The following publications relate to the prior art relating to the present invention. JP-A-60-171677 JP-A-3-194774

The technique described in Japanese Patent Application Laid-Open No. Sho 60-171677 (a method of processing defects in a storage medium) includes a map table for storing the use state of the storage medium in block units. When a defective block is detected, defective block information is stored in the map table, and then an unused block is searched. If this unused block is detected, I / O is performed from that block (unused block).
The O (Input / Output) processing is restarted. This corresponds to a defective block that occurs after the start of use of the magnetic disk device. However, in this technique, if an unused block is not detected, I / O processing is interrupted, and all I / O requests that need to secure a defective block thereafter become errors.

Further, in the technique described in the above-mentioned Japanese Patent Application Laid-Open No. 3-194774 (the defective block replacement processing method of the external storage device), the number of remaining blocks in the replacement area of the defective block is reduced to a predetermined value. At times, the fact is reported to the host system, thereby avoiding issuing a useless "REASSIGN BLOCK command" when the alternative area is used up. However, even with this technique, any I / O request to the corresponding disk device that needs to secure a defective block results in an error.

[0005] FIG. 8 shows an example of the configuration of a conventional technique (conventional defective block management system).

[0006]

The problem with the prior art described above is that, when the replacement area of the defective block in the disk device is exhausted, the I / O which needs to secure the defective block generated thereafter is needed. O requests are all errors.

[0007] The reason is that, since the alternative area is exhausted, an area (block) for data recording is recorded on the recording medium.
Cannot be secured, and data recording becomes impossible.

SUMMARY OF THE INVENTION In view of the above, it is an object of the present invention to provide a computer system including a plurality of disk devices each having a defective block alternative area on a recording medium, wherein one disk apparatus runs out of its own defective block alternative area. In this case, the I / O can be performed by using the replacement area of the defective block of another disk device in the computer system.
An object of the present invention is to provide a defective block management method capable of avoiding error completion of a request.

[0009]

According to the defective block management method of the present invention, an I / O request issued from a higher-level software is used in a computer system including a plurality of disk devices each having a defective block replacement area on a recording medium. I
Issue a replacement block allocation request issued from a certain disk controller to a disk controller other than the disk controller, and issue a defective block in the certain disk device. When a block in the replacement area in another disk device is used as an alternative to the disk control device corresponding to the disk device that provided the replacement area when the disk device having the defective block was replaced. A disk higher-level control unit that issues a substitute block release request
A defective block management table in the disk control device that holds the defective block position, defective block remapping destination information, and “information indicating the use state of the replacement area of the own disk device” for the own disk device; The defective block management table is updated so that unused blocks in the replacement area in the own disk device are allocated as being used during the block allocation processing, and the used blocks of the replacement area in the own disk apparatus are replaced during the replacement block release processing. The defective block management table is updated so as to be released and unused, and if a replacement block has been changed during I / O processing within the own disk, the defective block management table is updated based on the change. The management table updating unit in the disk control device that performs the A disk device output control unit issues an I / O request in determining the disk controller of the completion status of the I / O completion interrupt from the own disk device location,
Performs general control in the disk control device including notification of normal completion and error completion. When a request block of an I / O request from the disk higher-level control unit is registered in the defective block management table as a defective block, A disk control device that converts a request block into a remapping destination block, determines whether the remapping destination is a block in an alternative area in the disk device, and recognizes an I / O request issue target based on the determination. And a control unit.

[0010]

Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of a defective block management system according to the present invention.

The defective block management system according to the present embodiment includes a host software 10, a disk host controller 20, a disk controller 31 (disk controller identified by # 1), and a disk controller 32 (identified by # 2). Disk device 41 (a disk device identified by # 1) having a replacement area (substitution area for defective blocks) 411 and a disk device 42 (a disk identified by # 2) having a replacement area 421 Device). The disk control device can be realized as hardware or software.

The disk host control unit 20 has the following:
The following control is performed.

The I / O request issued from the host software 10 is sent to a disk controller (disk controller 31 or disk controller 32) corresponding to the disk device (disk device 41 or disk device 42) to be I / O. Issue.

A substitute block allocation request issued from a certain disk controller (for example, disk controller 31) is issued to a disk controller other than the disk controller (in the above example, disk controller 32).

A certain disk device (disk device 41)
Or 42) as a replacement for the defective block in another disk device (disk device 42 or 41).
When the disk unit 41 or 42 having the defective block is replaced when the block 21 or 411 is used, the disk control unit corresponding to the disk unit 42 or 41 that provided the replacement area 421 or 411 An alternative block release request is issued to 32 or 31.

The disk controller 31 (32) includes a controller 310 (320) and a defective block management table 311.
(321) and the management table update unit 312 (322)
And a disk device input / output control unit 313 (323).

The control section 310 (320) performs overall control in the disk control device 31 (32) (specific control contents relating to the present invention will be described later).

The defective block management table 311 (32
1) is its own disk device (corresponding disk device 41)
(42)) has management information on the defective block. That is, the “defective block position” (information indicating the position of the defective block) and the “defective block remapping destination information” (remapping destination of the replacement block of the defective block (the replacement block is Indicating the existing disk device and the position of the replacement block in the disk device), and “the replacement area 411 (421) of the own disk device”.
"Information indicating the use state of". " Here, the “information indicating the use state of the replacement area 411 (421) of the own disk device” is the information indicating the position of each block in the replacement area 411 (421) and whether each block is in use or not. It is assumed that the information indicates whether the information is used.

The management table update unit 312 (322)
The defective block management table 311 (321) is updated (update for realizing assignment and release of a substitute block to a defective block).

The disk device input / output control unit 313 (32
3) issues an I / O request to the disk device 41 (42) and performs a process related to an interrupt from the disk device 41 (42).

FIG. 2 shows the overall processing (I / O request from the host software 10 to the disk device 41 (disk device identified by # 1) when the I / O request is issued in the defective block management system of this embodiment. 9 is a flowchart showing a processing flow of (a process in which is issued). This processing includes an upper software I / O request issuance step 2-1, an upper disk controller I / O request issuance step 2-2, a request block defective block determination step 2-3, O processing ”success / failure determination step 2-4;
"Normal completion processing" execution step 2-5, alternative block allocation request issuing step 2-6, "unused alternative block allocation processing" success / failure determination step 2-7, and "error completion processing" execution step 2-8. "I / O in own disk"
Processing ”success / failure determination step 2-9 and“ error completion processing ”
Execution step 2-10, alternative block change presence / absence determination step 2-11, and "normal completion processing" execution step 2-
12, a table update request issuing step 2-13,
It comprises a "normal completion processing" execution step 2-14, a request block remapping destination block conversion step 2-15, a remapping determination step 2-16 in the own disk device, and an I / O request issuing step 2-17.

FIG. 3 is a flowchart showing the processing flow of "I / O processing in own disk". This processing includes an I / O request issuance step 3-1, a completion status receiving step 3-2, an I / O target block defective block corresponding determination step 3-3, a success recognition step 3-4, and "unused". "Alternate block allocation process" success / failure determination step 3-5;
It comprises a failure recognition step 3-6.

FIG. 4 is a flowchart showing the processing flow of the “unused alternative block allocation processing”. This process includes an unused block remaining determination step 4-1, a failure recognition step 4-2, and an unused block in use update step 4-4.
3, an I / O target block changing step 4-4, and a success recognition step 4-5.

FIG. 5 is a flowchart showing the processing flow of the "normal completion processing". This processing includes a disk upper control unit normal completion notification step 5-1, a higher software normal completion notification step 5-2, and an I / O processing end step 5-
3

FIG. 6 is a flowchart showing the processing flow of the "error completion processing". This processing includes a disk upper-level control unit error completion notification step 6-1, a higher-level software error completion notification step 6-2, and an I / O processing end step 6-3.

FIG. 7 is a flowchart showing the processing flow of the "alternate block release processing". This processing includes a disk device exchange step 7-1, a disk upper control unit sensing step 7-2, and an alternative block release request issuing step 7-3.
, Defective block management table check step 7-4
And an alternative block unused state return step 7-5.

Next, the operation of the defective block management system of this embodiment will be described with reference to FIGS.

(1) Overall operation when I / O request is issued

Here, the overall operation of the defective block management system of this embodiment when an I / O request to the disk device 41 (disk device identified by # 1) is issued from the upper software 10 will be described. (See FIG. 2).

The host software 10 sends an I / O request to the disk host control unit 20 (the target device of the request (I / O target)).
Issues an I / O request which is the disk device 41 (step 2-1).

Upon receiving the I / O request, the disk higher-level control unit 20 determines that the target device of the request is the disk device 41, and checks the disk controller 31 (disk device 41).
The I / O request is issued to the corresponding disk controller (step 2-2).

Upon receiving the I / O request, the disk controller 31 performs the following steps 2-3 to 2-6.

First, the control unit 310 refers to the defective block management table 311 and determines whether or not the requested block is registered as a defective block in the defective block management table 311 (step 2-3). In other words, the same “defective block position” as the position of the request block
Is determined in the defective block management table 311.

If it is determined in step 2-3 that "the required block is not registered as a defective block in the defective block management table 311 (the required block is not a defective block)", then "I / O in own disk" The control unit 310 determines whether the “in-disk I / O process” has succeeded or failed (step 2-4). The contents of the “in-disk I / O processing” for a certain disk controller will be described later.

If the "I / O processing in own disk" in step 2-4 succeeds, "normal completion processing" is performed (step 2-5). The contents of the "normal completion process" for a certain disk control device will be described later.

On the other hand, in step 2-4, "I
If the “/ O process” fails, the control unit 310 controls the disk control device 32 via the disk host control unit 20.
A substitute block allocation request is issued to (the disk control device identified by # 2) (step 2-6).

The disk controller 32 which has received the substitute block allocation request performs the processing shown in the following steps 2-7 to 2-14.

First, "unused alternative block allocation processing"
The control unit 320 determines whether the “unused alternative block allocation process” has succeeded or failed (step 2-7). The contents of the “unused alternative block allocation process” for a certain disk control device will be described later.

If the "unused substitute block allocation process" in step 2-7 has failed, "error completion process" is performed (step 2-8). The content of the “error completion process” for a certain disk control device will be described later.

On the other hand, if the “unused alternative block allocation processing” in step 2-7 is successful, the “in-disk I / O processing” is performed, and the “in-disk I / O processing” is performed by the control unit 320. Is successful or unsuccessful (step 2-9).

If the "I / O processing in own disk" in step 2-9 fails, "error completion processing" is performed (step 2-10).

On the other hand, in step 2-9, "I
If the "/ O process" succeeds, the "I
The control unit 320 determines whether or not the replacement block has been changed during the "/ O process" (step 2-11).
That is, "unused alternative block allocation processing" is executed in the "own disk I / O processing" (see step 3-5), and the I / O target block is changed in the "unused alternative block allocation processing". (Step 4-
4) has been performed. "

If it is determined in step 2-11 that "alternative block has not been changed", "normal completion processing" is immediately performed (step 2-12).

On the other hand, if it is determined in step 2-11 that "alternative block has been changed",
20 to the disk controller 31 (the disk controller that issued the alternative block allocation request in step 2-6) via the disk higher-level control unit 20 to update the table update request Defective block management table 311 in control device 31
Is issued (step 2-1).
3) Then, "normal completion processing" is performed (step 2-).
14).

The disk controller 31 determines in step 2-3
If it is determined that "the required block is registered as a defective block in the defective block management table 311 (the required block is a defective block)", the processing shown in the following steps 2-15 to 2-17 is performed. .

First, the control block 310 assigns the requested block to the remapping destination block (defective block management table 3).
The defective block in 11 (that is, the requested block)
(The block specified by the “defective block remapping destination information” corresponding to the “defective block position”) (the remapping destination block is set as a new request block) (step 2-15).

Further, the remapping destination block is controlled by the control unit 310 so that the replacement area 4 in the disk device 41 (the disk device corresponding to itself, that is, the own disk device).
11 (whether it is remapped to the replacement area 411 in the disk device 41) (step 2-16).

If it is determined in step 2-16 that "the remapping destination block exists in the replacement area 411 in the disk device 41", the process proceeds to step 2-4. The processing contents after step 2-4 are as described above.

On the other hand, in step 2-16, "the remapping destination block is set to the alternate area 42 in another disk device (a disk device other than the own disk device, ie, disk device 42)
1, the disk controller 32 (a disk controller corresponding to the disk device 42) via the disk higher-level controller 20 by the controller 310.
Issue an I / O request (step 2-17).

After receiving the I / O request, the disk controller 32 proceeds to the process of step 2-9. The processing contents after step 2-9 are as described above.

(2) Operation in “I / O processing in own disk”

The operation of the "I / O processing in own disk" for a certain disk controller (here, the disk controller 31) will be described below with reference to FIG.

Disk control unit 31 receiving I / O request
First, by the disk device input / output control unit 313,
Issue an I / O request for the I / O target block of the disk device 41 (own disk device) (step 3-
1).

Next, the disk device input / output control unit 313 receives a completion status indicated by an I / O completion interrupt returned as one of the hardware interrupts after completion of the actual I / O processing in the disk device 41 ( Step 3-2).

Further, the disk device input / output control unit 313
The completion status is checked to determine whether the I / O target block has become a defective block (step 3-3).

If it is determined in step 3-3 that "the I / O target block is not a defective block", the control unit 310 recognizes that the "in-disk I / O processing" is successful. (Step 3-4). Note that here, it is assumed that error completion does not occur except for the fact that the I / O target block has become a defective block (in practice, such error completion may occur, but in that case, the scope of the present invention) Is processed in the same manner as in the prior art by control other than the above.)

If it is determined in step 3-3 that "the I / O target block has become a defective block", "unused alternative block allocation processing" described later is performed, and "unused alternative block allocation processing" is performed. Is determined by the control unit 310 (step 3-
5).

If it is determined in step 3-5 that "the unused block allocation process has succeeded", the process returns to step 3-1 again (an I / O request for a new I / O target block). Is issued to the disk device 41).

When it is determined in step 3-5 that "unused substitute block allocation processing has failed", the control unit 310 recognizes that "internal disk I / O processing" has failed (step S5). 3-6).

(3) Operation in “unused alternative block allocation processing”

The operation of the "unused alternative block allocation process" for a certain disk controller (here, the disk controller 31) will be described with reference to FIG.

The disk control unit 31 refers to “information indicating the use state of the substitute area 411 of the own disk device” in the defective block management table 311 by the control unit 310, and reads the information in the disk device 41 (own disk device). It is determined whether an unused block remains in the substitute area 411 (step 4-1).

If it is determined in step 4-1 that “unused blocks remain in the substitute area 411”, the control unit 310 executes “unused substitute block allocation processing”.
(Step 4-2).

On the other hand, in step 4-1, the "alternative area 411"
Is determined, the management table updating unit 312 updates the defective block management table 311 so that one of the unused blocks is being used (step 4-3). ).

Further, the control unit 310 changes the I / O target block to the block used in step 4-3 (step 4-4).

When the processing of step 4-4 is completed,
The control unit 310 recognizes that the “unused alternative block allocation process” has succeeded (step 4-5).

(4) Operation in "normal completion processing"

The operation of the "normal completion process" for a certain disk controller (here, the disk controller 31) will be described with reference to FIG.

The disk controller 31 notifies the disk higher-level controller 20 of "normal completion" by the controller 310 (step 5-1).

The disk higher-level control unit 20 that has received the notification of step 5-1 notifies the "normal completion" to the higher-level software 10 (step 5-2). As a result, I / O requests for I / O requests issued from higher
The O processing ends (step 5-3).

(5) Operation in “Error Completion Processing”

The operation of the "error completion process" for a certain disk controller (here, the disk controller 31) will be described with reference to FIG.

The disk controller 31 notifies the disk host controller 20 of "error completion" by the controller 310 (step 6-1).

Upon receiving the notification of step 6-1, the disk host controller 20 notifies the host software 10 of the "error completion" (step 6-2). by this,
I for I / O request issued from upper software
The / O processing ends (step 6-3).

(6) Operation in "alternative block release processing"

The disk device 41 has its own substitute area 41
Disk unit 42 because the blocks in 1 have been used up.
If the block in the replacement area 421 in the disk device 41 is used as a replacement block, the replacement block in the replacement area 421 in the disk device 42 used for the disk device 41 at the timing when the disk device 41 is replaced is replaced. Need to be released.

The operation in the "alternate block release processing" according to the defective block management method of this embodiment in such a case will be described below with reference to FIG.

When the disk device 41 (the disk device identified by # 1) is replaced (step 7-1), the disk higher-level control section 20 senses that fact (step 7-1).
2) Issue a substitute block release request to the disk controller 32 (step 7-3).

Upon receiving the substitute block release request, the disk control unit 32 checks the defective block management table 321 by the management table updating unit 322 (step 7-4), and substitutes the substitute block used for the disk unit 41. The defective block management table 321 is updated so that (the block in the replacement area 421) is returned to the unused state (step 7-5).

In this embodiment, as shown in FIG.
The description has been made on the assumption that the number of disk devices and disk control devices is two. However, the number of disk devices and disk control devices need only be plural, and is not limited to two. If the number is 3 or more, the disk host controller 20 selects one disk controller other than the disk controller when a certain disk controller issues an alternative block allocation request. Issue the substitute block allocation request to the disk controller (the selected disk controller).

As a selection algorithm in this case, for example, “a disk controller having the largest number of unused blocks obtained as a response to the inquiry to each disk controller about the number of unused blocks in an alternate area in each disk device” Is selected as a disk control device corresponding to "." When such a selection algorithm is employed, the alternative area of each disk device can be efficiently used as the entire computer system.

[0083]

As described above, according to the present invention,
In a computer system including a plurality of disk devices having a defective block replacement area on a recording medium, a disk higher-level control unit for managing all disk devices may be provided above each disk control device, By providing a defective block management table having defective block remapping destination information capable of indicating the position of the alternative area, for example, a certain disk controller can issue an alternative block allocation request to another disk controller. Even when the disk device runs out of its own substitute area, it is possible to avoid returning an I / O request to the disk device with an error. This means that an I / O target block of an I / O request issued to a disk device that has used up its own substitute area can be replaced with another block in the computer system until the disk device is replaced. This is because it is possible to substitute a block in a replacement area in the disk device.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a defective block management system according to the present invention.

FIG. 2 is a flowchart showing a processing flow of an overall process when an I / O request of the defective block management method shown in FIG. 1 is issued.

FIG. 3 is a flowchart showing a processing flow of “I / O processing in own disk” in the defective block management method shown in FIG. 1;

FIG. 4 is a flowchart showing a processing flow of “unused alternative block allocation processing” in the defective block management method shown in FIG. 1;

FIG. 5 is a flowchart showing a processing flow of “normal completion processing” in the defective block management method shown in FIG. 1;

FIG. 6 is a flowchart showing a processing flow of “error completion processing” in the defective block management method shown in FIG. 1;

FIG. 7 is a flowchart showing a processing flow of “alternate block release processing” in the defective block management method shown in FIG. 1;

FIG. 8 is a block diagram showing a configuration of an example of a conventional defective block management method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Upper software 20 Disk upper control part 31, 32 Disk control device 41, 42 Disk device 310, 320 Control part 311, 321 Defective block management table 312, 322 Management table update part 313, 323 Disk device input / output control part 411, 421 Alternative area

Claims (3)

(57) [Claims] In a computer system including a plurality of disk devices provided with a replacement area for a defective block on a recording medium, a disk control corresponding to a disk device to which an I / O request issued from higher-level software is subjected to I / O. Issue a replacement block allocation request issued from a certain disk control device to a disk control device other than the disk control device, and as a replacement for a defective block in a certain disk device, a replacement area in another disk device. A disk higher-level control unit that issues a substitute block release request to a disk control device corresponding to a disk device that provided an alternative area when a disk device having the defective block was replaced when a block was used. And defective block position and defective block remapping for own disk device Block management table in the disk controller that holds the information and "information indicating the use status of the replacement area of the own disk device", and the unused area of the replacement area in the own disk device when the unused replacement block is allocated The defective block management table is updated so as to allocate the block as being used, and the defective block management table is released so that the used block of the substitute area in the own disk device is released and used when the substitute block is released. And a management table updating unit in the disk control device for updating the defective block management table based on the change if the replacement block has been changed during the I / O processing in the own disk; At the time of internal I / O processing, I / O
A disk device input / output control unit in the disk control device that issues a request and determines the completion status of an I / O completion interrupt from the own disk device, and general control in the disk control device including notification of normal completion and error completion If the requested block of the I / O request from the disk higher-level control unit is registered as a defective block in the defective block management table, the request block is converted into a remapping destination block, and the remapping destination is changed to the local disk. A control unit in a disk control device for determining whether or not the block is a block in an alternative area in the device and recognizing an I / O request issuance target based on the determination. 2. An algorithm for selecting an issue destination disk control device when issuing a substitute block allocation request issued from a certain disk control device to a disk control device other than the disk control device is based on a “replacement area in each disk device”. Query the disk controller for the number of unused blocks of the disk controller, and select the disk controller corresponding to the disk device with the largest number of unused blocks obtained as a response. The defective block management system according to claim 1, wherein 3. The “information indicating the use state of the replacement area of the own disk device” is information indicating the position of each block in the replacement area and information indicating whether each block is in use or unused. 3. The defective block management system according to claim 1, further comprising a certain defective block management table.