JP2003108316A - Disk array device and hot spare control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide, in a disk array system including a hot spare, a disk array device that effectively makes use of the hot spare when the system is in a normal operating state to improve the access speed in data read and to speed up the array restoration processing if a fault occurs. SOLUTION: A RAID controller 15, in normal operation, sequentially, selects given block data in each write block data for disk devices (Disk 1, Disk 2, Disk 3, Disk 4) 16a to 16d and writes the given block data in a disk device 16n (Disk S) of the hot spare. In addition, the controller, in normal operation, makes concurrent read access to all of Disk 1, Disk 2, Disk 3, Disk 4 and Disk S.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is equipped with or can be equipped with a hot spare (standby disk device).
For example, RAID (redundant arrays of inexpensive d
The present invention relates to a disk array device and a hot spare control method suitable for use when constructing an isks system.

[0002]

2. Description of the Related Art In a disk array device such as a RAID system, in order to realize a quick recovery function at the time of failure in consideration of the failure of each of a plurality of disk devices constituting an array. The system configuration is such that a standby disk device called a hot spare that can always be replaced can be installed.

An example of the operation of the conventional hot spare at this time will be described with reference to FIG.

In the example shown in FIG. 5, four disk devices D are used.
A RAID 5 (RAID level 5) system is constructed by using the desk 1, the disc 2, the disc 3, and the disc 4, and a hot spare (hot spare disk) is prepared as a standby disk device.

In the figure, (a) is a normal operating state, and disk devices Disk1, Disk2, Disk3, Di
The sk4 operates according to the RAID5 predetermined striping write rule. For example, at a certain timing, Dis
Stripe data is written to k1, 2, 3 and
The redundant data (parity) is written in k4. At the next timing, the stripe data is written to Disks 2, 3 and 4, and the redundant data (parity) is written to Disk 1. Under this normal operating state, the hot spare (HotSpareDisk) is in a standby state in which it can operate.

In the state shown in (a), when a failure (failure) occurs in the disk device Disk2 as shown in (b), a failed disk device Di as shown in (c).
sk2 is separated from the array, and the remaining disk device Di
The degenerate operation is performed by sk1, Disk3, and Disk4, and disk devices Disk1, Disk3, and Disk are used for the hot spare (HotSpareDisk).
The contents of the failed disk device Disk2 are restored from 4. When this restoration is completed, as shown in (d), the disk devices Disk1, Disk3, Disk4, and the hot spare (HotSpareDisk) are assigned RAID5.
Works.

As described above, in this type of system configuration including a hot spare, under normal conditions,
The hot spare is always (immediately) ready for operation to prepare for the failure of the disk device that constitutes the array.
Therefore, the hot spare is in the energized state even though it does not operate under the normal condition and the motor is driven, so that the power is wasted.

As described above, conventionally, in a disk array system equipped with or capable of being equipped with a hot spare, when the system is in a normal operating state, the hot spare is placed in a useless standby state.

[0009]

As described above, in the conventional disk array system in which the hot spare is prepared as the standby disk device, when the system is in the normal operating state, the hot spare is put in the useless standby state. Had been

The present invention has been made in view of the above circumstances,
In a disk array system equipped with or capable of being equipped with a hot spare, when the system is operating normally, the hot spare is effectively used to improve the access speed of data read and the speed of array restoration processing in the event of a failure. It is an object of the present invention to provide a disk array device and a hot spare control method for realizing the same.

[0011]

According to the present invention, a hot spare is made effective by writing a partial write data of an array to a hot spare in duplicate during normal operation and simultaneously reading the disk devices including the hot spare at the time of reading. This is utilized to improve the performance at the time of read with high access frequency. Furthermore, because some data of the array is stored in the hot spare, it is not necessary to restore the above-mentioned stored data in the event of an array disk failure, and the time required for the restoration process can be shortened. And

That is, the disk array device of the present invention is a disk array device comprising a plurality of disk devices forming an array according to a predetermined RAID level and a standby disk device, wherein the standby disk device and the plurality of disks are provided. A control unit that simultaneously operates the device and selects a predetermined data block from the data blocks to be written to each disk device every time the data block is written to the plurality of disk devices, and sequentially writes the selected data block to the standby disk device. It is characterized by having.

Further, the disk array device of the present invention is a disk array device comprising a mirror disk device composed of at least two disk devices and a standby disk device, wherein the standby disk device and the mirror disk device are provided. It is characterized by comprising control means for operating simultaneously and writing the write data block to the mirror disk device also to the standby disk device.

Further, the disk array device of the present invention is a disk array device for writing data with a predetermined striping over a plurality of disk devices, the hot spare disk device, the hot spare disk device and the plurality of disk devices. Control to operate simultaneously and select a predetermined data block from the data blocks to be written to each disk device every time the data blocks forming the striping are written to the plurality of disk devices, and sequentially write to the standby disk device It is characterized by comprising means.

Further, the disk array device of the present invention is characterized in that, at the time of reading a data block, different data blocks are simultaneously read from a plurality of disk devices and a standby disk device.

The disk array device of the present invention, when one of the plurality of disk devices fails, disconnects the failed disk device and restores the stored contents of the failed disk device from the contents stored in the remaining disk devices. When storing in the standby disk device, restoration control is performed so that the stored content corresponding to the failed disk device stored in the standby disk at the time of writing is not restored.

With this configuration, the hot spare can be effectively used to increase the data read speed, and the RAID 1 system can eliminate the data recovery time of the failed disk. Also, R
In the AID3 system, by writing data and parity in the hot spare, the recovery time from a failure can be improved. Further, in the RAID5 system, by writing only valid data other than parity to the hot spare, it is possible to improve the read performance and the recovery time from a failure.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an example of the configuration of a computer system to which a disk array device to which the present invention is applied.

In the system shown in FIG. 1, a CPU 11, a system memory 12, a display controller 13, a keyboard controller 14, a RAID controller 15 and the like are connected to a PCI bus 1.

The CPU 11 controls the entire system and executes processing according to various application programs including an operating system and utility programs stored in the system memory 12.

The system memory 12 is a memory device serving as a main memory of the computer system, and stores an application program including an operating system and utilities that describe the operation of the CPU 11 and various processing data used for executing these. To do.

The display controller 13 controls the output of the user interface of this computer system, and is a video RA drawn by the CPU 11.
Display data in M131 is displayed on CRT132 or LCD133.
The display is controlled on the display device such as.

The keyboard controller 14 controls the input of the user interface of the computer system, and converts the operation of the input device such as the keyboard 141 and the mouse 142 into data and transfers it to the CPU 11.

The RAID controller 15 controls a plurality of disk devices 16a to 16n forming a disk array, which is an external storage of this computer system.
Data is exchanged with the plurality of disk devices 16a to 16n forming a disk array via the SCSI bus 2.

2 to 4 are operation explanatory diagrams in the first to third embodiments of the present invention which are executed under the control of the RAID controller 15, respectively.
3 is an operation explanatory view in the first embodiment of the present invention, FIG.
4 is an operation explanatory view in the second embodiment of the present invention, FIG.
[Fig. 8] is an operation explanatory diagram according to the third embodiment of the present invention.

Here, the operation of the embodiment of the present invention will be described with reference to the drawings.

First, the operation of the first embodiment of the present invention will be described with reference to FIG.

The first embodiment shown in FIG. 2 is a RAID
A disk array device of RAID1 is configured by the three disk devices 16a, 16b, 16n placed under the control of the controller 15. Of these, two disk devices 16a and 16b constitute mirroring (RAID 1), and the remaining one disk device 16n is a hot spare (standby disk device). This hot spare may be optional. Here, the disk device 1
6a is Disk1, disk device 16b is Disk2,
The disk device 16n to be a hot spare is shown as DiskS. Further, the disk on the logical address (logical disk) viewed from the CPU 11 is DiskL, and the logical data blocks to be handled are Block1, Block2, and
It is indicated by….

In the first embodiment, the DiskS, which is the hot spare during normal operation, is
The same write data block (Block as Block2)
1, Block2, ...) is written. Here, Dis
Since k1 and Disk2 have a RAID1 (mirror) configuration, the stored contents of each Disk1 = Disk2 = DiskS are the same.

In this state, for example, D forming a mirror
When a failure occurs in the disk2, the RAID controller 15 disconnects the failed Disk2,
The RAID 1 (mirror) is newly configured by 1 and the DiskS to enable continuous operation, and therefore the restoration time is not required.

Further, during normal operation, the RAID controller 15 simultaneously performs read access to all of Disk1, Disk2, and DiskS, so that, for example, in data read processing, for example, data blocks "Block0-Block0-
When reading "Block5", "Bl" is read from Disk1.
"ock0", Disk2 to "Block1", Disk
Read "Block2" from kS at the same time, then D
“Block3” from desk1 and “Bl from Disk2”
"lock4" and "Block5" from DiskS are controlled to be read at the same time. This allows existing RAID1
Data can be read at a higher speed than the two mirrors of Disk 1 and Disk 2 according to the configuration.

Next, the operation of the second embodiment of the present invention will be described with reference to FIG.

The second embodiment shown in FIG. 3 is a RAID
By the five disk devices 16a, 16b, 16c, 16d, 16n placed under the control of the controller 15, R
A disk array device of AID3 is constructed. Of these, three disk devices 16a, 16b and 16c are striped to form a data drive, one disk device 16d forms a parity dedicated drive, and the remaining one disk device 16n is a hot spare (standby disk). Device). This hot spare may be optional. Here, the disk device 16a is set to Di.
sk1, the disk device 16b is shown as Disk2, the disk device 16c is shown as Disk3, the disk device 16d that is a parity dedicated drive is shown as DiskP, and the disk device 16n that is a hot spare is shown as DiskS. Further, a disk (logical disk) on a logical address viewed from the CPU 11 is represented by DiskL, and logical data blocks to be handled are represented by Block1, Block2, ....

In the second embodiment, the RAID controller 15 is a hot spare during normal operation.
to diskS, Disk1, Disk2, Disk3, D
A predetermined data block is selected from the data blocks written in the iskP, and the operation is controlled so that the data blocks are sequentially stored. Alternatively, as shown in FIG. 3, copying is performed in a fixed order in units of Disk. For example, RAID controller 1
5 indicates that the first data block of DiskS is
Redundant data P (0-2) same as kP is written, Block3 same as Disk1 is written in the next data block, Block7 same as Disk2 is written in the next data block, and next data block is written in the next data block. Block 11, which is the same as that of Disk 3, is written. Thereafter, similarly, writing is performed as shown in FIG.

In this state, for example, a defect occurs in Disk 2 forming a stripe and D becomes a hot spare.
Restore Disk2 to diskS (in this case, Disk
When restoring the data of Disk2 from DiskP1, Disk3, Disk3), the RAID controller 15 uses DiskS
In the writing process, the Dis before
Block data Block7, Block stored in k2
Since each data of ck19, ... Has been written, it operates so as not to be restored. As a result, the restoration time of the Disk 2 can be shortened compared to the conventional case where all data blocks are restored. The restoration time is not limited to the failure of Disk2, and the failure time of Disk1, 3, P can be similarly shortened.

Further, during normal operation, the RAID controller 15 operates as follows: Disk1, Disk2, Disk3, Disk.
For example, the data block “Block0
~ Block3 "is read from Disk1 to" B
"lock0", Disk2 to "Block1", Di
“Block2” from sk3, “Blo” from DiskS
ck3 ”is controlled to be read at the same time,
It is possible to read using the hot spare disk device DiskS and use the existing RAID5 configuration
Data can be read at a higher speed than reading using 1, Disk 2, Disk 3.

Next, the operation of the third embodiment of the present invention will be described with reference to FIG.

The third embodiment shown in FIG. 4 is a RAID
By the five disk devices 16a, 16b, 16c, 16d, 16n placed under the control of the controller 15, R
A disk array device of AID5 is constructed. Of these, four disk devices 16a, 16b, 16c, 16
d constitutes writing of striping data with parity (writing of redundant data (parity) is not fixed to one disk device and is distributed and written to the disk devices 16a to 16d), and the remaining one disk device 16
n is a hot spare (standby disk device). This hot spare may be optional. Here, the disk device 16a is Disk1, and the disk device 1 is
6b is Disk2, disk device 16c is Disk3,
The disk device 16d is shown as Disk 4, and the hot spare disk device 16n is shown as DiskS. Further, a disk (logical disk) on a logical address viewed from the CPU 11 is represented by DiskL, and logical data blocks to be handled are represented by Block1, Block2, ....

In the third embodiment, the RAID controller 15 is a hot spare during normal operation.
to diskS, Disk1, Disk2, Disk3, D
A predetermined data block is selected from the data blocks to be written to isk4, and the operation is controlled so that the data blocks are sequentially stored. Alternatively, as shown in FIG. 4, copying is performed in a fixed order in units of Disk. For example, the same Block1 as the Disk2 is written in the first data block of the DiskS, and the same B as the Disk1 is written in the next data block.
lock4 is written, and D is written in the next data block.
Block7, which is the same as disk4, is written, and Block10, which is the same as Disk3, is written in the next data block. Thereafter, similarly, writing is performed as shown in FIG. In this state, for example, a failure occurs in Disk2 and DiskS, which is a hot spare,
Restore 2 (in this case, Disk1, 3, and Disk
When restoring the data of Disk2 from 4), RAID
The controller 15 has already written the data blocks Block1 and Block1 to the DiskS before the failure of the Disk2.
Since each data of 3, ... Has already been written, it is controlled not to be restored. As a result, the restoration time can be shortened as compared with the conventional case where all data blocks are restored. It should be noted that the present invention is not limited to the failure of Disk2,
Similarly, with respect to the failures 1, 3, and 4, the restoration time can be shortened.

Further, during normal operation, the RAID controller 15 operates as follows: Disk1, Disk2, Disk3, Disk.
4, Read access to all of DiskS at the same time,
As the data read processing, for example, when reading the data blocks “Block0 to Block4”, Di
"Block0" from sk1 and "Blo from Disk2"
ck1 ”, Disk3 to“ Block2 ”, Disk
4 to "Block3", DiskS to "Block"
4 ”at the same time, it is possible to perform reading using the hot spare disk device DiskS, and to use the existing RAID5 configuration, Disk1,
Data can be read at a higher speed than reading using Disk2, Disk3, and Disk4.

[0042]

As described in detail above, according to the present invention, in a disk array system equipped with a hot spare, when the system is in a normal operating state, the hot spare can be effectively utilized to improve the access speed of data read. And the speed of array restoration processing when a failure occurs can be achieved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration example of a computer system to which a disk array device to which the present invention is applied.

FIG. 2 is an operation explanatory diagram according to the first embodiment of the present invention.

FIG. 3 is an operation explanatory diagram according to the second embodiment of the present invention.

FIG. 4 is an operation explanatory diagram according to the third embodiment of the present invention.

FIG. 5 is a diagram for explaining an operation example of a conventional hot spare.

[Explanation of symbols]

11 ... CPU 15 ... RAID controller 16a to 16n ... Disk device

Claims (12)

[Claims] 1. A disk array device comprising a plurality of disk devices forming an array according to a predetermined RAID level and a standby disk device, wherein the standby disk device and the plurality of disk devices are simultaneously operated, and A disk comprising control means for selecting a predetermined data block from data blocks to be written to each disk device and writing the data sequentially to the standby disk device each time a data block is written to a plurality of disk devices. Array device. 2. A disk array device comprising a mirror disk device composed of at least two disk devices and a standby disk device, wherein the standby disk device and the mirror disk device are simultaneously operated, and the mirror disk A disk array device comprising control means for writing a data block to be written to the device to the standby disk device. 3. The disk array device according to claim 2, further comprising control means for reading different data blocks from the mirror disk device and the standby disk device at the same time when reading the data block. 4. A disk array device for writing data with a predetermined striping across a plurality of disk devices, wherein the hot spare disk device, the hot spare disk device and the plurality of disk devices are operated at the same time, Each time a data block forming the striping is written to the disk device, a predetermined data block is selected from the data blocks to be written to each disk device, and the control means is sequentially written to the standby disk device. Disk array device. 5. The disk array device according to claim 4, further comprising control means for reading different data blocks from the plurality of disk devices and the standby disk device at the same time when reading the data blocks. 6. When one of the plurality of disk devices fails, the failed disk device is disconnected, the stored contents of the failed disk device are restored from the contents stored in the remaining disk devices, and the standby disk device is restored. 5. The disk array device according to claim 4, wherein, when the data is stored, restoration control is performed so as not to restore the storage content corresponding to the failed disk device stored in the standby disk at the time of writing. 7. A hot spare control method for a disk array device comprising a plurality of disk devices forming an array according to a predetermined RAID level and a standby disk device serving as a hot spare, wherein the standby disk device and the plurality of standby disk devices are provided. Each time a data block is written to the plurality of disk devices, a predetermined data block is selected from the data blocks to be written to each disk device and sequentially written to the standby disk device. Hot spare control method characterized by. 8. A hot spare control method for a disk array device comprising a mirror disk device composed of at least two disk devices and a standby disk device, wherein the standby disk device and the mirror disk device are operated simultaneously. A hot spare control method, wherein a write data block to the mirror disk device is also written to the standby disk device. 9. The hot spare control method according to claim 8, wherein different data blocks are simultaneously read from the mirror disk device and the standby disk device at the time of reading the data block. 10. A hot spare control method for a disk array device comprising a hot spare disk device for performing data writing with predetermined striping across a plurality of disk devices, wherein the hot spare disk device and the plurality of disk devices are operated simultaneously. Each time a data block forming the striping is written to the plurality of disk devices, a predetermined data block is selected from the data blocks written to each disk device and sequentially written to the standby disk device. And hot spare control method. 11. When reading the data block,
11. The hot spare control method according to claim 10, wherein different data blocks are simultaneously read from the plurality of disk devices and the standby disk device. 12. When one of the plurality of disk devices fails, the failed disk device is disconnected, the stored contents of the failed disk device are restored from the contents stored in the remaining disk devices, and the standby disk device is restored. 11. The hot spare control method according to claim 10, wherein when storing, the stored contents of the failed disk device stored in the standby disk at the time of writing are not restored.