JP2009003789A - Power failure handling method for disk device, and disk device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a power failure handling method for a disk device that saves battery power, by backing up data with hardware unnecessary for the data backup stopped in a power failure. <P>SOLUTION: The power failure handling method for a disk device is for use with an uninterruptible power supply, a drive enclosure comprising first and second disk units and a drive control unit, and a disk control device comprising a disk cache memory and a controller unit. The disk control device is configured to back up data in the disk cache memory to the second disk unit as stopping the rotation of the first disk unit on receiving a power failure detection notification from the uninterruptive power unit. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a power failure processing method for a disk device, and more specifically, a power failure processing for stopping power consumption of a battery of an uninterruptible power supply by stopping a disk unit unnecessary for data backup at the time of a power failure and performing data backup of a disk cache memory It is about the method.

In recent years, the amount of data stored and processed by computers has increased significantly, and the system has become larger. In a large-scale system, for example, a collective disk device using a plurality of disk units using a RAID (Redundant Arrays of Inexpensive Disks) technique is widely used.
In these systems, an uninterruptible power supply (UPS) that supplies power to the disk device for a while when a power failure occurs due to some cause in the commercial power supply is provided. In this uninterruptible power supply, the occurrence of a power outage is detected and notified, and a battery built in the uninterruptible power supply supplies power to the disk device instead of commercial power.
Upon receiving the notification of the power failure, the disk device performs data backup that saves the data stored in the disk cache memory included in the disk device to a predetermined disk unit. Then, the data saved in the disk unit when the power is restored is taken out and written (stored) in the disk unit that should be stored.
FIG. 4 shows a configuration example of a conventional disk device 100. The disk device 100 includes a disk control device 200 and a drive enclosure 300.
The disk control device 200 includes controller units 210 and 220 that control the drive enclosure 300 (more specifically, the disk units 310 and 320). In general, the controller units 210 and 220 have a disk cache memory and are configured in two units in order to increase redundancy. Normally, two units are operated, and when a failure occurs in the controller unit, the controller unit in which the failure has occurred is separated and operated by one unit. Further, controller power supply units 230 and 240 for supplying power to the controller units 210 and 220 are incorporated. “# 0” and “# 1” in the figure are numbering for distinguishing the same type.
The drive enclosure 300 distributes signals from the m system disk units 310, the n user disk units 320, and the controller units 210 and 220 to the disk units, and selects signals from the disk units to select the controller unit 210. , 220 are configured by disk control units 330 and 340 for performing control to be sent to 220. The system disk unit 310 is a disk unit dedicated to the system, and the user cannot store data in the system disk unit 310. The user disk unit 320 is a disk unit opened to the user and stores user data. The disk control units 330 and 340 have two units for the same reason as the controller units 210 and 211. When one disk control unit fails in normal operation, one unit is operated by the remaining disk control units. It becomes. In addition, drive enclosure power supply units 350 and 360 are incorporated to supply power to the disk control units 330 and 340 and the motor that rotates the disk unit.
The disk device 100 configured as described above includes the uninterruptible power supply 400, which switches from the commercial power supply in the event of a power failure and supplies power to the disk control device 200 and the disk enclosure 300 from the battery included in the uninterruptible power supply 400. Then, the data backup described above is performed. In general, the disk device 100 is supplied with power from a commercial power supply via the breakers 110 and 120, and the uninterruptible power supply 400 is supplied with power for charging from the commercial power supply.
The disk device 100 is coupled to a host computer (not shown) through a channel, and takes data to be written to the disk device 100 into the disk cache memory via the channel. Then, the data taken into the disk cache memory is written into the disk unit. Since the time for writing data to the disk cache memory is faster than the time for writing to the disk unit, the disk cache memory is interposed between the host computer and the disk unit, so that writing (storage) to the disk device as seen from the host computer side is possible. The processing can be completed in a short time (when the disk cache memory acquires the write data, the host computer is notified of the completion of writing to the disk device, and at this point, the host computer is released from this processing. Thereafter, data is written from the disk cache memory to the disk unit as described above). The disk cache memory also temporarily stores data read from the disk unit and sends it to the host computer via the channel. This is because when there is a data read request from the host computer to the disk device, the disk cache memory is first examined, and if there is the requested data, there is no need to read from the disk unit. A method of processing a disk device against a power failure of a commercial power source has been proposed.

  In the first proposal, when a power failure is detected, the program executed by the computer is stopped, the program information is stored in the main memory, the contents of the main memory are saved in the nonvolatile storage device, and an indicator indicating the occurrence of the power failure is displayed. Set it. When the indicator is set at the time of recovery from a power failure, the stored contents are recovered from the nonvolatile storage device to the main storage unit, information on the program is set in the processing unit, and the process is resumed (Patent Document 1).

The second proposal is to check the voltage of the battery at the time of a power failure, and store the data stored in the disk cache and not written to the disk device when the remaining power amount becomes equal to or less than the power amount required for saving. (Patent Document 2).
Japanese Patent Application Laid-Open No. 07-21091 JP 09-330277 A

  As described above, there has been a remarkable increase in the amount of data processing in recent times, and the amount of data stored in the disk device is constantly increasing. For this reason, an expansion of a disk device or a replacement with a large-capacity disk device has been carried out, and accordingly, the power capacity of the battery of the uninterruptible power supply installed at the same time is also increasing. The increase in power capacity has led to an increase in the size of the uninterruptible power supply, and securing the installation space has become an issue.

  The above-mentioned first proposal can reduce the battery size because the hardware necessary for the information saving operation in the event of a power failure can be localized only in the processing unit, main storage unit, and nonvolatile storage device. . However, the first proposal does not consider the data backup of the disk cache, and does not suppress the power consumption related to the disk device.

  In the second proposal, data backup is not performed immediately in the event of a power failure, so that it is not necessary to perform data backup when power is restored after a short power failure. However, it is considered that the same power supply as before is required from the battery until immediately before the data backup is performed.

  The present invention cuts power to hardware unnecessary for backup of disk cache data at the time of a power failure, and reduces power consumption by reducing the battery capacity. It aims to save space.

The disk device power failure processing method and the disk device of the present invention are configured as follows.
(1) First invention A power failure processing method for a disk device according to a first invention comprises an uninterruptible power supply device in which a disk device has a built-in battery to notify the detection of a power failure, and a disk unit comprising first and second power units. A drive enclosure configured with a drive control unit that controls distribution of signals to the disk unit, and a disk control device configured with a disk cache memory and a controller unit that controls access to the drive control unit. The control device receives a power failure detection notification from the uninterruptible power supply, stops the rotation of the first disk unit, and backs up the data in the disk cache memory to the second disk unit. .

The first disk unit is a user disk unit that can write user data, for example, and the second disk unit is a system disk unit provided exclusively for the system, for example.
(2) Second invention The power failure processing method for the disk device according to the second invention is such that the disk control device in the power failure processing method according to the first invention stops the output of the circuit that controls the first disk unit of the drive control unit. It is characterized by that.
(3) Third invention A power failure processing method for a disk device according to a third invention is a redundant configuration in which the disk control device in the power failure processing method according to the first and second inventions has two drive control units and two controller units. The disk controller receives a power failure detection notification from the uninterruptible power supply and checks the remaining capacity of the uninterruptible power supply battery. If the remaining capacity is less than the specified value, the drive control unit and controller unit The power supply on one side is stopped, and data is backed up by the drive control unit and controller unit on the remaining one side.
(4) Fourth invention A disk device according to a fourth invention comprises an uninterruptible power supply device incorporating a battery and notifying the detection of a power failure, first and second disk units, and the disk unit. Upon receiving a power failure detection notification from the drive enclosure and the uninterruptible power supply that controls the signal distribution, the first disk unit stops rotating and the data in the built-in disk cache memory is And a disk control device for backing up to two disk units.
(5) Fifth Invention The disk device of the fifth invention is such that the disk control device in the power failure processing method of the fourth invention stops the output of the circuit that controls the first disk unit of the drive control unit. It is a feature.

  As described above, according to the present invention, the following effects can be obtained.

  According to the first invention, power saving can be achieved by stopping the rotation of the disk unit that is not required for data backup, and the battery built in the uninterruptible power supply can be reduced in size. Therefore, it is possible to provide a power failure processing method that can realize space saving.

  According to the second aspect of the invention, in addition to stopping the rotation of the disk unit that is not necessary for data backup, the output of the drive control unit of the circuit that controls the disk unit is stopped, so that the power consumption of the battery can be further reduced and the power can be cut off. Space saving of the power supply device can be realized.

  According to the third aspect of the invention, the drive control unit of the redundantly configured drive enclosure and the controller unit of the disk control device are operated on one side, so that it is possible to save battery power.

  According to the fourth invention, it is possible to provide a disk device having the same effects as those of the first invention.

  According to the fifth aspect of the invention, it is possible to provide a disk device that has the same effect as the second aspect of the invention.

  FIG. 1 shows an example of a disk device 101 during a power failure according to the present invention. FIG. 1 corresponds to the configuration example of the disk device shown in FIG. 4 and shows a state in which hardware unnecessary for data backup is turned off in the event of a power failure (the one indicated by a dotted line is powered by a battery) Shows no hardware). That is, the one side drive control unit 340 configured redundantly with the n user disks 320 of the drive enclosure 300 and the one side controller unit 220 configured redundantly of the disk control device 200 are indicated by dotted lines. The rotation of the user disk 320 is stopped by a rotation stop command (spin down command) from the controller of the controller unit 210, for example. As a result, it is not necessary to supply power to the motor that rotationally drives n disk units, and power saving can be achieved. The drive control unit 340 normally operates together with the drive control unit 330 in a redundant configuration. However, at the time of a power failure, only the drive control unit 330 is moved and the drive control unit 340 is stopped. As a result, it is possible to save power required for one drive control unit. Similarly to the drive control unit, the controller unit 220 normally operates together with the controller unit 210 in a redundant configuration. However, the controller unit 220 can be stopped to supply power to one controller unit. Power saving.

  Although not shown in FIG. 1, the circuit of the part that controls the user disk 320 in the drive control unit 330 is also turned off. It is possible to save power in this portion.

  Although FIG. 1 shows an example in which hardware power and functions unnecessary for data backup are stopped, the present invention will be described in more detail with reference to FIGS. FIG. 2 shows the controller unit 210 of the disk controller 200 that is the main component of the present invention in more detail in the same configuration as FIG. The controller unit 210 includes a CPU 211, a cache memory 212, an input / output control unit 213, a channel adapter 214, and a device adapter 215 (here, only hardware related to the present invention is shown. The controller unit 220 is also shown. It is the same as the controller unit 210).

  The CPU 211 controls the entire disk control device. For example, the CPU 211 obtains a power failure notification from the uninterruptible power supply 410 and issues a spin-down command to the drive enclosure 300 or performs data backup.

  The cache memory 212 is the disk cache memory described above, and data sent in response to a write request from the host computer 500 is temporarily stored in the cache memory 212. Thereafter, the data stored in the cache memory 212 is written into the disk unit.

  The input / output control unit 213 is connected to the uninterruptible power supply 410 and receives a power failure detection notification from the uninterruptible power supply 410. Further, the CPU 211 requests the uninterruptible power supply 410 to measure the battery voltage, and acquires the result.

  The channel adapter 214 interfaces with the host computer 500, which is a host device of the disk device 101, and transmits / receives write / read data.

  The device adapter 215 interfaces with the drive enclosure 300, which is a lower device of the disk device 101, and transmits / receives write data from the cache memory 212 or read data to the cache memory 212.

  The uninterruptible power supply 410 has a built-in battery and notifies the controller units 210 and 220 when a power failure is detected, measures the battery voltage based on the request of the controller units 210 and 220, and notifies the result. In addition, as shown in the figure, power is supplied from the battery during a power failure to the controller unit 210 of the disk control device 200, the system disk 310 of the drive enclosure 300, and the drive control unit 330.

  Next, the processing flow at the time of a power failure of the controller unit 210 will be described with reference to FIG. First, when the controller unit 210 receives notification from the uninterruptible power supply 410 that a power failure has been detected, the controller unit 210 issues a spin-down command for requesting the user disk unit to stop rotating, to the drive enclosure 300. Here, the drive enclosure 300 is equipped with m system disk units and n user disk units, and the motor that rotationally drives the n user disk units is stopped by a spin-down command (S100, S110).

  Subsequently, for the drive control units 330 and 340 of the drive enclosure 300, n partial circuits that control the user disk units in the circuits of the drive control units 330 and 340 are turned off. Specifically, for example, a switch circuit is configured in the output section of the control circuit, and a program that turns off the output when this switch operates on the corresponding disk unit circuit based on the command is stored in the firmware. Can be realized by mounting them in the drive control units 330 and 340 (S120).

  Next, an instruction to measure the voltage of the battery built in the uninterruptible power supply 410 is issued to the uninterruptible power supply 410, and the measurement result is acquired. The uninterruptible power supply 410 measures the battery voltage at the request of the controller unit 210 and notifies the result. The controller unit 210 checks whether or not the acquired battery voltage is less than or equal to a predetermined value set as a threshold value. If the battery voltage exceeds the predetermined value, the controller unit 210 determines that the battery has sufficient remaining power capacity and stores it in the disk cache memory. Data is backed up to the system disk unit 310. If the acquired battery voltage is equal to or lower than the predetermined value, the controller unit 220 and the disk control unit 340 are turned off, and then data backup is started (S130-S160).

  According to the above, when a power failure occurs, power supply to hardware unnecessary for data backup can be cut off and data backup can be performed with the minimum amount of hardware, so the battery of the uninterruptible power supply 410 has a small capacity Will be enough. For this reason, a battery can be reduced in size and the installation space of the uninterruptible power supply 410 can also be suppressed.

In addition to the above example, the following additional notes are disclosed.
(Appendix 1)
An uninterruptible power supply with built-in battery and notification of power failure detection;
A drive enclosure composed of a first and second disk unit and a drive control unit for controlling signal distribution to the disk unit;
A power failure processing method for a disk device comprising a disk cache memory and a disk control device comprising a controller unit for controlling access to the drive control unit,
The disk control device receives a power failure detection notification from the uninterruptible power supply, stops the rotation of the first disk unit, and backs up the data in the disk cache memory to the second disk unit. A power failure processing method for a disk device.
(Appendix 2)
2. The disk device power failure processing method according to claim 1, wherein the disk control device stops output of a circuit that controls the first disk unit of the drive control unit.
(Appendix 3)
The disk device operates the drive control unit and the controller unit in two redundant configurations, respectively.
The disk controller receives a power failure detection notification from the uninterruptible power supply and checks the remaining capacity of the battery of the uninterruptible power supply, and when the value of the remaining capacity is equal to or less than a predetermined value, 3. The power failure processing method for a disk device according to appendix 1 or appendix 2, wherein power supply to one side of the controller unit is stopped and data backup is performed with the drive control unit and the controller unit on the remaining one side.
(Appendix 4)
An uninterruptible power supply with built-in battery and notification of power failure detection;
A drive enclosure composed of a first and second disk unit and a drive control unit for controlling signal distribution to the disk unit;
A disk control device that receives a power failure detection notification from the uninterruptible power supply, stops the rotation of the first disk unit, and backs up the data of the built-in disk cache memory to the second disk unit. A disk device characterized by the above.
(Appendix 5)
The disk device according to appendix 4, wherein the disk control device stops the output of a circuit that controls the first disk unit of the drive control unit.
(Appendix 6)
4. The disk device according to any one of appendices 1 to 3, wherein the first disk unit is a disk unit that is open for user data writing, and the second disk unit is a system disk unit. Power failure processing method.
(Appendix 7)
4. The disk device power failure processing method according to appendix 3, wherein the remaining capacity of the battery is obtained based on a voltage value of a battery built in the uninterruptible power supply.

It is an example of the disk apparatus at the time of the power failure by this invention. It is a connection example of the configuration of the controller unit and other hardware. It is a power failure processing flow of the controller unit. 2 is a configuration example of a conventional disk device.

Explanation of symbols

100 Disk Device 101 Disk Device 110 Power Breaker 120 Power Breaker 200 Disk Controller 210 Controller Unit 211 CPU
212 Cache Memory 213 Input / Output Control Unit 214 Channel Adapter 215 Device Adapter 220 Controller Unit 230 Controller Power Supply Unit 240 Controller Power Supply Unit 300 Drive Enclosure 310 System Disk 320 User Disk 330 Drive Control Unit 340 Drive Control Unit 350 Drive Enclosure Power Supply Unit 360 Power supply unit for drive enclosure 400 Uninterruptible power supply 410 Uninterruptible power supply 500 Host computer

Claims (5)

An uninterruptible power supply with built-in battery and notification of power failure detection;
A drive enclosure composed of a first and second disk unit and a drive control unit for controlling signal distribution to the disk unit;
A power failure processing method for a disk device comprising a disk cache memory and a disk control device comprising a controller unit for controlling access to the drive control unit,
The disk control device receives a power failure detection notification from the uninterruptible power supply, stops the rotation of the first disk unit, and backs up the data in the disk cache memory to the second disk unit. A power failure processing method for a disk device. 2. The power failure processing method for a disk device according to claim 1, wherein the disk control device stops output of a circuit that controls the first disk unit of the drive control unit. 3. The disk device operates the drive control unit and the controller unit in two redundant configurations, respectively.
The disk controller receives a power failure detection notification from the uninterruptible power supply and checks the remaining capacity of the battery of the uninterruptible power supply, and when the value of the remaining capacity is equal to or less than a predetermined value, 3. The power failure processing method for a disk device according to claim 1, wherein power supply to one side of the controller unit is stopped, and data backup is performed by the drive control unit and the controller unit on the other side. 4. . An uninterruptible power supply with built-in battery and notification of power failure detection;
A drive enclosure composed of a first and second disk unit and a drive control unit for controlling signal distribution to the disk unit;
A disk control device that receives a power failure detection notification from the uninterruptible power supply, stops the rotation of the first disk unit, and backs up data in the built-in disk cache memory to the second disk unit. A disk device characterized by the above. The disk device according to claim 4, wherein the disk control device stops output of a circuit that controls the first disk unit of the drive control unit.

Images