JP4751241B2 - Disk array device - Google Patents

Description

  The present invention relates to power saving of computer devices, peripheral devices, and other electronic devices, and more particularly to energy saving of a disk array device connected to a computer device.

  There is a constant demand for higher functionality and lower prices in electronic devices. Recently, environmental issues have become more important and energy saving has been demanded.

  A disk array device is a peripheral device of a computer device and is used as an external storage device for storing computer data. When a disk array device is equipped with a plurality of magnetic disk devices, and the disk array device and the computer system as the host device are functioning, the plurality of magnetic disk devices mounted on the disk array device are usually used. Everything is working.

  In these magnetic disk devices, the access target magnetic disk device must be in operation so that control information and data can be transmitted or received when an access from a host device occurs. On the other hand, when there is no access from the host device or the control device for a predetermined time, it is not necessary to keep the magnetic disk device “in operation”.

  Here, “in operation” means that when a control information or data is input from a control device or a host device to the magnetic disk device, the magnetic disk device is in a state of immediately responding. . This does not include responding after elapse of the time that the so-called magnetic disk device is in the sleep mode and the magnetic disk medium is spun up from the stationary state to the normal rotational speed. It does not include responding after the magnetic disk device is in the power saving mode and canceling the power saving mode.

  Further, the access means whether a command directed to itself is issued as viewed from the magnetic disk device, and whether exchange of information intended for itself is requested. Usually, in response to a command for acquiring or storing information of the host device, it corresponds to a signal being issued from the electronic circuit on the host device side including the drive interface to the target magnetic disk device.

  In general, a magnetic disk device is turned on when an access from a host device is received, a predetermined electronic circuit is turned on, a magnetic recording medium is rotated, and a response (information transmission or reception) is made to access. There is a control method in which the power of a motor for rotating a predetermined electronic circuit or magnetic recording medium is turned off stepwise after completion. In other words, the magnetic disk device has various power saving modes built-in, and has a function of selecting a predetermined power saving mode and automatically shifting to that mode mainly according to the access frequency from the host device or the control device. Have.

  In the conventional disk array device, the mounted magnetic disk device or the magnetic disk device set so as to be recognizable by the host device is controlled so as to be in operation when the disk array device is activated. After the disk array device is activated, the magnetic disk device mounted thereon is in operation until the power is turned off. When an operation for turning off the power of the disk array device is performed, a sequence for turning off the magnetic disk device from “in operation” is triggered by this, and then the power of the disk array device is cut off.

  A magnetic disk device that is not set to be recognizable by the host device and a magnetic disk device that is mounted for spare use are set to be usable from the host device, or a spare magnetic disk device is used. Each time the power is turned on. Once the magnetic disk device was in operation, it was not individually turned off until the disk array device was turned off. For this reason, when the number of magnetic disk devices increases, for example, in a disk array device that connects several hundred magnetic disk devices, a power saving measure becomes essential.

  In the disk array device, when there are few accesses from the host device, it is not necessary to keep all the magnetic disk devices in operation. For this reason, it is considered that power saving can be achieved by applying the above control method. However, when the access is received, time is required for the magnetic disk device to be “in operation”, and the performance of the entire disk array device is significantly reduced. In order to suppress the performance degradation, select and execute the power saving mode with a short recovery time until “in operation”, use a magnetic disk device with a short time to spin-up, or replace the magnetic disk device. In addition, access to the memory holding the data must be considered.

  Further, in a disk array device that does not handle a set of magnetic disk devices (physical units) as one logical unit, it is not possible to realize power saving of the disk array device simply by diverting the existing power saving mode of the magnetic disk device. .

  For each set of magnetic disk device groups (physical unit groups), when there is no access for a predetermined time from the host device or the control device, one of a plurality of power saving modes is selected and the set of magnetic disk device groups Is the selected power saving mode. This power saving mode includes a mode in which the rotation of the magnetic recording medium is stopped.

  When one set of magnetic disk device groups (physical unit groups) supports two or more logical units, all of the two or more logical units are not accessed from the host device or control device for a predetermined time. Then, one of the plurality of power saving modes is selected, and the set of magnetic disk device groups is set to the selected power saving mode.

  When one logical unit is defined beyond a set of magnetic disk device groups (physical unit groups), a higher-level device or a control device is used for all the magnetic disk devices corresponding to the one logical unit. When there is no access for a predetermined time, one of a plurality of power saving modes is selected, and the set of magnetic disk device groups is set to the selected power saving mode.

In order to execute such control, in the present invention, the disk array device is configured to access a plurality of disk devices and an access transmitted from a higher-level device to the first logical unit among the plurality of logical units. And a control unit that controls to store data in the first plurality of disk devices corresponding to the first logical unit of the plurality of disk devices, the control unit configured to The operation state of the first plurality of disk devices corresponding to the first logical unit is in a first operation state that does not produce a power saving effect, and is managed in advance by the control unit without being accessed. When the predetermined time has elapsed, the operating states of the first plurality of disk devices corresponding to the first logical unit are The first logical unit corresponding to the first logical unit is controlled so as to be changed to one second operating state among the plurality of second operating states having a power saving effect in a plurality of disk device units. In the case where the first plurality of disk devices corresponding to the first logical unit are diagnosed when the operation state of the plurality of disk devices is the second operation state, the first logical unit Diagnosis is performed on all of the first plurality of disk devices corresponding to the above. Here, diagnosis refers to checking whether or not the disk device is in a usable state. For example, an online verify command is executed, and it is confirmed whether or not the command ends normally.
Further, in the present invention, a method for controlling a disk array device including a plurality of disk devices, the step of receiving an access transmitted from a host device to a first logical unit among a plurality of logical units; Controlling to store data in a first plurality of disk devices corresponding to the first logical unit of the plurality of disk devices in accordance with the received access; and the first logical unit The first plurality of disk devices configuring the first disk device are in a first operation state in which no power saving effect is achieved, and the first logical unit is not accessed and is managed in advance by the control unit If the time has elapsed, the operations of the first plurality of disk devices corresponding to the first logical unit Controlling the state to change to a second operating state of one of the plurality of second operating states having a power saving effect in units of the first plurality of disk devices, and the first logical When diagnosing the first plurality of disk devices corresponding to the first logical unit when the operation state of the first plurality of disk devices corresponding to the unit is the second operation state And a step of performing diagnosis on all of the first plurality of disk devices corresponding to the first logical unit.

  The present invention can suppress the power consumption of the magnetic disk device mounted in the disk array device, and further suppresses a significant performance degradation of the disk array device because it targets a magnetic disk device that is not accessed by a host device. However, energy saving is possible. Further, by performing diagnosis on a magnetic disk device whose power is cut off, it is possible to confirm the reliability of the magnetic disk device whose power is often cut off.

  More specifically, for example, in a single magnetic disk device that requires 20 W of power at the time of rotation startup, the electronic circuit consumes 5.5 W during read / write and 3.5 W during read / write idle. When the spindle of the magnetic disk device is idling, power consumption is 4W to 5W. In this case, if the power saving mode for stopping the function of the electronic circuit is selected, the power consumption of at least 3.5 W is suppressed, and if the rotation of the spindle is stopped, the power consumption of 4 W to 5 W is additionally suppressed. It will be.

  In fact, since power is saved in units of RANKs, if the disk array has a 5-column configuration, if one RANK electronic circuit is in the power saving mode, power consumption of about 18 W can be achieved. In addition, power consumption of 20 to 25 W is suppressed. These numbers may vary depending on how the disk array device is used. For example, the power saving effect is increased by the change in daytime and nighttime access demand in the continuous operation for 24 hours and the execution time of the power saving mode.

  Embodiments of the disk array device of the present invention will be described below.

  FIG. 1 shows an example of the internal configuration of the disk array device 110 connected to the host device 101.

  The host device 101 controls reading and writing of information and issues a command to the disk array device 110 to realize reading and writing of information.

  The disk array device 110 includes a host interface circuit 120, a control unit 160, a data transfer circuit 170, a drive interface circuit 180, a magnetic disk device 190, a spare magnetic disk device 191, an operation panel 130, a LAN control unit 140, and RS232C control. The unit 150 is configured.

  The control unit 160 is realized by a microprocessor and control firmware. It comprises a main control unit 161 that controls the entire apparatus and a time measuring mechanism 163 that manages time. Various management tables are placed in the memory 162 on the control unit 160.

  For the magnetic disk device 190 and the spare magnetic disk device 191, if a small-sized magnetic disk device, which is a general-purpose part, is used, the effect of reducing the manufacturing cost is great. The magnetic disk devices 190 are arranged in an array so as to have a RAID (Redundant Array Inexpensive Disks) configuration (FIG. 1).

  The RAID group is composed of one column (200, 201) or a plurality of columns (202). The magnetic disk device 190 constituting the RAID group introduces the concept of “RANK” in order to make it accessible from the host device 101, and sets a logical unit for each RAID group configuration. The storage area of the magnetic disk device is divided within each same rank (area division).

  In general, in a disk array, for example, user data is divided into data of an appropriate size (striping) for a plurality of magnetic disk devices, and is distributed and stored in each magnetic disk device. The disk array preferably stores and accesses data evenly to its own magnetic disk device. For this purpose, the concept of RANK is made to correspond to a configuration composed of a plurality of logical magnetic disk devices. Therefore, a set of magnetic disk devices can correspond to a plurality of or less than one logical disk device.

  The association between the RANK and the logical unit is determined by setting the configuration information of the disk array device 110. nRANK = 1 logical unit, 1RANK = n logical unit, or mRANK = n logical unit. In FIG. 1, LU0 (200) is set to LU0 (210, logical unit number 0), RANK1 (201) is set to LU1 (211) and LU2 (212), and RANKn (202) is set to LU3 (213) and LU4 (214). It is.

  The spare magnetic disk device 191 is used as a substitute magnetic disk device when a failure occurs in the magnetic disk device 190 constituting the RAID group. Specifically, the diagnosis is executed to detect the failure, and the failed magnetic disk device 191 and the spare magnetic disk device 191 are replaced. Although FIG. 1 shows only one spare magnetic disk device, the present invention is not limited to this.

  FIG. 2 shows various management table structures on the memory 162.

  The magnetic disk device management table 250 manages information for each set logical unit. The set logical unit is managed by registering its number in the set LUN. In the case of the configuration of RANKn (202), since it has a multi-column configuration, the branch number of the division is registered and managed here using the management LUN. In addition, the time when access to the logical unit is received from the host apparatus 101 is registered as the access time. The position of the magnetic disk device 190 in the set logical unit is registered in the magnetic disk device position. When a plurality of logical units are set in the same RANK as shown in RANKn (202), the set other logical unit numbers are registered in the related LUN. The time when the magnetic disk device 190 is set to the power saving mode or the time when the power of the magnetic disk device 190 is shut off is registered as the power saving start time.

  The power saving waiting time 260 registers the time until the magnetic disk device 190 enters the power saving mode or the power is turned off after the last access from the host device 101 is received.

  The diagnosis start time 270 is the time from when the magnetic disk device 190 enters the power saving mode or the like until the diagnosis of the magnetic disk device 190 starts, or the time to execute the diagnosis of the magnetic disk device 190 that enters the power saving mode or the like. sign up.

  Hereinafter, a power saving method according to the present invention will be described with reference to flowcharts (FIGS. 3, 4 and 5).

  The host device 101 executes an application and stores necessary information in the disk array device 110. Necessary information is extracted or written by read or write access to the disk array device, and necessary functions are activated.

  When the disk array device 110 is started up, the necessary magnetic disk device 190 is turned on so that it can respond to access from the host device.

  Since the spare magnetic disk device 191 is not normally accessed from the host device 101, the power is turned off after the power is turned on and it is confirmed that it operates normally. Instead of shutting off the power source of the spare magnetic disk device 191, the standby magnetic disk device 191 may be put on standby in a power saving mode.

  The disk array device 110 turns on the power of the magnetic disk device 190, confirms the normal operation of the magnetic disk device 190, and then sets the set LUN, management LUN, access time (normality of the magnetic disk device 190) in the storage device management table 250. At the time of confirmation), the magnetic disk device position, and the associated LUN are registered (FIG. 3). The management LUN registers information for dividing and managing the logical unit for each column when the RANKn (202) is composed of a plurality of magnetic disk devices 190. The LU 4 (214) is registered in the management LUN as LU 4-1 and LU 4-2. As shown in RANK 1 (201), when a plurality of logical units are set in the same RANK, Register the associated LUN. LU2 (212) is registered in the related LUN of LU1 (211), and LU1 (211) is registered in the related LUN of LU2 (212).

  After registering such information, the power-off process shown in FIG. 4 and the diagnostic process shown in FIG. 5 are started. Instead of the power-off process, a process for shifting to the power saving mode may be used. Hereinafter, since the power-off process has a large power saving effect, the embodiment will be mainly described. However, the power saving mode may be selected and executed instead of the power-off.

  The access time and the power saving start time (power off time) are registered with reference to the current time from the time measuring mechanism 163 in the control unit 160.

  When access from the host device 101 is received, the power saving start time of the magnetic disk device 190 belonging to the logical unit to be accessed is determined from the magnetic disk device management table 250, and if the power saving start time is not registered, it continues. Access. When the power saving start time is registered, the power of the magnetic disk device is off or in the power saving mode. Therefore, the access is executed after the power of the magnetic disk device 190 registered at the magnetic disk device position in the magnetic disk device management table 250 is turned on or the power saving mode is canceled and the power saving start time is cleared. After the execution of access is completed, the access time is updated to the current time.

  In the power-off process of FIG. 4, the access time of the magnetic disk device management table 250 is monitored for each set logical unit. The time when the logical unit that is no longer accessed from the host device is compared with the current time registered as the access time, and when the difference exceeds the power saving waiting time 260, the magnetic disk device 190 belonging to the target logical unit is Judging from the position of the magnetic disk device in the magnetic disk device management table 250, the power supply of the magnetic disk device 190 registered at the magnetic disk device position is shut off or is set to the power saving mode. The time when the power is shut off is registered as the power saving start time (FIG. 2).

  When the power saving waiting time 260 is exceeded, the LU1 (211) determines whether to turn on or off the power (depending on the power saving mode) according to the state of the LU2 (212) of the associated LUN registered. If LU2 (212) does not exceed the power saving waiting time, it is determined that the power saving start time of the magnetic disk device management table 250 is not registered and the magnetic disk device 190 is being used in LU2 (212). Thus, the power supply of the magnetic disk device is not shut off or is not set in the power saving mode. If LU2 (212) has exceeded the power saving waiting time, the power saving start time of the magnetic disk device management table 250 has already been registered, and the power saving waiting time 260 has been exceeded together with LU1 (211). The power of the magnetic disk device 190 is shut off. The time when the power is shut off is registered as the power saving start time.

  When the power saving waiting time 260 is exceeded, the LUN 4 (214) determines that the power is turned off in the registered management LUN. The LUN 4 (214) has a single logical unit configuration, but since it is composed of a plurality of rows of magnetic disk devices 190, it is divided and managed by the management LUN. The first column is registered as LUN4-1 and the second column is registered as LUN4-2. LUN 4-1 is registered with LUN 3 (213) in the related LUN, and when the power saving start time of LUN 3 (213) is not registered, it is in use by LUN 3 (213). The power of the eye magnetic disk device 190 is not turned off. Since the LUN 4-2 has no associated LUN, the power of the magnetic disk device 190 in the second column is turned off. The time when the power is turned off is registered at the power saving start time. In this way, the LUN 4 enables a part of the magnetic disk device group constituting one logical unit to be turned off or shifted to the power saving mode.

  In the diagnosis processing of FIG. 5, the power saving start time of the magnetic disk device management table 250 is monitored for each set logical unit. The time registered in the power saving start time is compared with the current time, and when the difference exceeds the diagnosis start time 270, the magnetic disk device 190 belonging to the target logical unit is replaced with the magnetic disk in the magnetic disk device management table 250. Judgment is made from the device position, and diagnosis of the magnetic disk device 190 registered at the magnetic disk device position is executed. After the diagnosis is completed, the power saving mode is started.

  When executing diagnosis, priority is given to access processing from the host device 101. When access is received, diagnosis is stopped, access is executed, and diagnosis is resumed after the access processing is completed.

  Diagnosis of the magnetic disk device 190 is performed by registering the time of the 24-hour clock at the diagnosis start time 270, so that the diagnosis starts at the registered time, not the elapsed time since the power is turned off. Further, diagnosis may be started at an elapsed time after the power is turned off.

  When executing diagnosis, priority is given to access processing from the host device 101. When access is received, diagnosis is stopped, access is executed, and diagnosis is resumed after the access processing is completed. The diagnosis is executed simultaneously for each logical unit registered in the magnetic disk device management table 250. When the logical unit is composed of a plurality of columns, the processing is executed in units of management LUNs that are managed in units of columns. The diagnosis of the maximum number of magnetic disk devices 190 in the column configuration is performed simultaneously. In the case of the configuration of FIG. 1, the diagnosis of five magnetic disk devices is simultaneously performed in each RANK including two magnetic disk devices (not shown). Here, for example, if four magnetic disk devices are mounted in one RANK, four diagnostics are performed simultaneously. This is to check before using the logical magnetic disk device as the RANK disk array.

  Although it is a special mode, the present invention can also be applied to a disk array having only one magnetic disk device in one RANK. In this case, one set of magnetic disk device group is composed of one magnetic disk device.

  The spare magnetic disk device 191 is diagnosed when a certain time has elapsed. When the time of the 24-hour clock is registered in the diagnosis start time 270, it may be performed when the registered time is reached.

  The power saving start time 260 and the diagnosis start time can be changed from the host command, the operation panel 130, the LAN port 140, and the RS232C port. When receiving a change instruction, the information in the memory 162 is updated. The change by the host command is performed by a vendor unique command instructing the change of the information. The change by the operation panel 130 is performed by displaying a change setting menu of these information on the operation panel and inputting a value to be changed by the operator. The change by the LAN port and the RS232C port is performed by specifying a specific interface in the LAN and the RS232C, embedding information to be changed in the interface, and sending the information from a device connected to the LAN and the RS232C. The changed information is written into the memory 162 when the change instruction is given, and becomes valid when the writing is completed.

1 is an overall configuration diagram of a disk array device according to an embodiment of the present invention. It is a figure of the management table concerning one Example of this invention. 3 is a flowchart showing a power-on operation of a magnetic disk device according to one embodiment of the present invention. 3 is a flowchart of power-off operation of a magnetic disk device according to one embodiment of the present invention. 5 is a flowchart of a power supply diagnosis operation for a magnetic disk device according to an embodiment of the present invention.

Explanation of symbols

101 ... Host device, 110 ... Disk array device,
120 ... Host interface circuit, 130 ... Operation panel,
140 ... LAN (local area network) port,
150 ... RS232C port, 160 ... control unit,
161: main control unit, 162: memory,
163 ... Timekeeping mechanism, 170 ... Data transfer circuit,
180 ... drive interface circuit, 190 ... magnetic disk drive,
191 ... Spare magnetic disk device, 200 ... RANK0,
201 ... RANK1, 202 ... RANK2,
210 ... LU (logical unit) 0, 211 ... LU (logical unit) 1, 212 ... LU (logical unit) 2, 213 ... LU (logical unit) 3, 214 ... LU (logical unit) 4,
250 ... Magnetic disk unit management table,
260 ... Power saving waiting time 270 ... Diagnosis start time.

Claims (6)

A disk array device,
Multiple disk units;
The first plurality of disk devices corresponding to the first logical unit of the plurality of disk devices in response to an access transmitted from the host device to the first logical unit of the plurality of logical units And a control unit that controls to store data in
With
The controller is
The operating state of the first plurality of disk devices corresponding to the first logical unit is in a first operating state that does not provide a power saving effect, and the first logical unit is not accessed by the control unit. When a predetermined time that is managed in advance elapses, the operation state of the first plurality of disk devices corresponding to the first logical unit is determined for each of the first plurality of disk devices. Control to change to one of the plurality of second operating states to be played,
Diagnosis of the first plurality of disk devices corresponding to the first logical unit when the operation state of the first plurality of disk devices corresponding to the first logical unit is the second operation state When performing the diagnosis, the diagnosis is performed on all of the first plurality of disk devices corresponding to the first logical unit.
A disk array device characterized by that. The disk array device according to claim 1,
An operation state of the second plurality of disk devices corresponding to a second logical unit of the plurality of logical units is in the first operation state, and the second logical unit is not accessed and the predetermined logical unit is not accessed. When the time has elapsed, the operation state of the second plurality of disk devices corresponding to the second logical unit is set to the second plurality of disk device units in the second operation state. Control to change to one second operating state,
When the operation state of the second plurality of disk devices corresponding to the second logical unit is the second operation state, diagnosis of the second plurality of disk devices corresponding to the second logical unit is performed. If so, the diagnosis is performed on all of the second plurality of disk devices constituting the second logical unit.
A disk array device characterized by that. The disk array device according to claim 1 or 2,
The controller is
Independently of the change in the operating state of the first plurality of disk devices corresponding to the first logical unit, the second plurality of disk devices corresponding to the second logical unit among the plurality of disk devices. Control to change the operating state
A disk array device characterized by that. A method of controlling a disk array device comprising a plurality of disk devices,
Receiving an access transmitted from a host device to a first logical unit of a plurality of logical units;
Controlling to store data in a first plurality of disk devices corresponding to the first logical unit of the plurality of disk devices in response to the received access;
The operation state of the first plurality of disk devices constituting the first logical unit is in a first operation state that does not provide a power saving effect, and the first logical unit is not accessed by the control unit. When a predetermined time that is managed in advance elapses, the operation state of the first plurality of disk devices corresponding to the first logical unit is determined for each of the first plurality of disk devices. Controlling to change to one of the plurality of second operating states to be played,
Diagnosis of the first plurality of disk devices corresponding to the first logical unit when the operation state of the first plurality of disk devices corresponding to the first logical unit is the second operation state Performing a diagnosis on all of the first plurality of disk devices corresponding to the first logical unit; and
A method for controlling a disk array device, comprising: The method of controlling a disk array device according to claim 4,
An operation state of the second plurality of disk devices corresponding to a second logical unit of the plurality of logical units is in the first operation state, and the second logical unit is not accessed and the predetermined logical unit is not accessed. When the time has elapsed, the operation states of the second plurality of disk devices corresponding to the second logical unit are changed to the second plurality of disk device units in the second plurality of disk device units. Control to change to one of the second operating state,
When diagnosing the second plurality of disk devices corresponding to the second logical unit when the operation state of the disk device corresponding to the second logical unit is the second operation state Execute diagnosis for all of the second plurality of disk devices corresponding to the second logical unit
A method for controlling a disk array device. A method for controlling a disk array device according to any one of claims 4 and 5,
Independently of the change in the operating state of the first plurality of disk devices corresponding to the first logical unit, the second plurality of disk devices corresponding to the second logical unit among the plurality of disk devices. Control to change the operating state
A method for controlling a disk array device.

Images