Storage system and power control method in a storage system

Abstract

To reduce the power consumption of the drives, the system divides the drives into those that are powered on and those that are powered off. [Solution] A storage system having multiple drives and processors that are accessed in response to I / O requests from a host determines which drives will have their power supply turned on and which drives will have their power supply turned off, based on the amount of data stored on the multiple drives. The storage system moves all the data stored on the drives to the drives to be powered off, and after moving all the data stored on the drives to be powered off, it turns off the power supply to the drives to be powered off.

Description

Technical Field

[0001] The present invention relates to a storage system and a power control method in the storage system.

Background Art

[0002] In recent years, with the development of IoT and AI, the amount of data stored and utilized in storage systems has been increasing. Along with this, the storage system has increased in capacity by increasing the number of drives for storing data such as SSDs and HDDs, and the power consumption has increased. Therefore, the increase in power consumption and the increase in electricity charges have become problems.

[0003] For example, Patent Document 1 discloses the following technology in order to reduce the power consumption of a storage system. That is, "The calculation unit calculates a first power consumption obtained by adding the power consumption of an unused first storage device to the total power consumption of a plurality of storage devices in operation, and adds the power consumption of an unused second storage device to the total power consumption of the plurality of storage devices in operation, and subtracts the power consumption of a storage device capable of moving data to the second storage device from among the plurality of storage devices in operation to obtain a second power consumption. The determination unit determines to activate the first storage device if the second power consumption is greater than or equal to the first power consumption, and determines to activate the second storage device if the second power consumption is not greater than or equal to the first power consumption" (see paragraph 0007 of cited reference 1).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, the conventional technology described above assumes that multiple drives are pre-divided into drives that are powered on and in operation, and drives that are powered off and unused. When an unused drive is powered on and added to the active drives, the power consumption of the drive is reduced. In other words, it does not divide multiple drives into drives that are powered on and drives that are powered off in order to reduce the power consumption of the drives.

[0006] This invention has been made in view of the above circumstances, and aims to reduce the power consumption of the drives by dividing multiple drives into drives that are powered on and drives that are powered off. [Means for solving the problem]

[0007] To achieve the above objective, the present invention, in one embodiment, provides a storage system having a plurality of drives and a processor that are accessed in response to I / O requests from a host, wherein the processor determines, based on the amount of data stored in the plurality of drives, which drives will have their power supply turned on and which drives will have their power supply turned off, moves all data stored in the power-saving drives to the normal drives, and after moving all data stored in the power-saving drives to the normal drives, turns off the power supply to the power-saving drives. [Effects of the Invention]

[0008] According to the present invention, for example, multiple drives can be divided into drives that are powered on and drives that are powered off in order to reduce the power consumption of the drives. [Brief explanation of the drawing]

[0009] [Figure 1] An explanatory diagram illustrating the overview of Embodiment 1. [Figure 2] A diagram showing the configuration of the storage system according to Embodiment 1. [Figure 3]A diagram showing the configuration of the drive information table according to Embodiment 1. [Figure 4] A diagram showing the configuration of the device information table according to Embodiment 1. [Figure 5] A diagram showing the configuration of the block information table according to Embodiment 1. [Figure 6] A diagram showing the configuration of the RAID group information table according to Embodiment 1. [Figure 7] A flowchart illustrating the drive power-off process according to Embodiment 1. [Figure 8] A flowchart illustrating the power saving target drive calculation process according to Embodiment 1. [Figure 9] A flowchart illustrating the calculation process for the minimum number of units according to Embodiment 1. [Figure 10] A flowchart illustrating the data storage count calculation process according to Embodiment 1. [Figure 11] A flowchart showing the performance limit check process according to Embodiment 1. [Figure 12] A flowchart illustrating the process for calculating the number of unused drives according to Embodiment 1. [Figure 13] A flowchart illustrating the drive group determination process according to Embodiment 1. [Figure 14] A flowchart illustrating the candidate drive extraction process according to Embodiment 1. [Figure 15] A flowchart illustrating the drive selection process according to Embodiment 1. [Figure 16] A flowchart illustrating the inactive data access process when the drive power is turned off, according to Embodiment 1. [Figure 17] A flowchart illustrating the inactive data addition and storage process according to Embodiment 1. [Figure 18] A flowchart illustrating the inactive data determination process according to Embodiment 1. [Figure 19] An explanatory diagram illustrating the overview of Embodiment 2. [Figure 20] A diagram showing the configuration of the storage system according to Embodiment 2. [Figure 21]Flowchart showing the drive group determination process according to Embodiment 2. [Figure 22] Flowchart showing the power-saving drive box accommodation drive determination process according to Embodiment 2. [Figure 23] Flowchart showing the candidate drive extraction process according to Embodiment 2. [Figure 24] Flowchart showing the drive selection process according to Embodiment 2. [Figure 25] Flowchart showing the drive power-off process according to Embodiment 3. [Figure 26] Flowchart showing the drive power-off process according to Embodiment 4.

Mode for Carrying Out the Invention

[0010] In the following description, an "interface device" may be one or more communication interface devices. The one or more communication interface devices may be one or more of the same type of communication interface devices (for example, one or more NICs (Network Interface Cards)) or two or more different types of communication interface devices (for example, a NIC and an HBA (Host Bus Adapter)).

[0011] Also, in the following description, a "memory" is one or more memory devices which are an example of one or more storage devices, and typically may be a main memory device. At least one of the memory devices in the memory may be a volatile memory device or a non-volatile memory device.

[0012] Also, in the following description, a "drive" is a persistent storage device. The persistent storage device is typically a non-volatile storage device (for example, an auxiliary storage device), and specifically, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), or an NVMe (Non-Volatile Memory Express) drive.

[0013] Furthermore, in the following explanation, "processor" may refer to one or more processor devices. At least one processor device may typically be a microprocessor device such as a CPU (Central Processing Unit), but may also be other types of processor devices such as a GPU (Graphics Processing Unit). At least one processor device may be single-core or multi-core. At least one processor device may be a processor core. At least one processor device may be a broad-sense processor device such as a hardware circuit that performs some or all of the processing (e.g., FPGA (Field-Programmable Gate Array), CPLD (Complex Programmable Logic Device), or ASIC (Application Specific Integrated Circuit)).

[0014] Furthermore, in the following explanation, the process may be described using "program" as the subject. A program, when executed by a processor, performs defined processes using memory devices and / or interface devices as appropriate. For this reason, the subject of the process may be the processor (or a device such as a controller having that processor). A program may be installed from a program source into a device such as a computer. The program source may be, for example, a program distribution server or a computer-readable (e.g., non-temporary) recording medium. Also, in the following explanation, two or more programs may be implemented as a single program, or one program may be implemented as two or more programs.

[0015] Furthermore, in the following explanation, we may use expressions such as "xxx table" to describe information from which an output is obtained for a given input. However, this information can be data of any structure (for example, structured data or unstructured data), or it can be a neural network that generates an output for a given input, a learning model such as a genetic algorithm or a random forest. Therefore, "xxx table" can be called "xxx information." Also, in the following explanation, the structure of each table is just an example, and one table may be divided into two or more tables, or all or part of two or more tables may be a single table.

[0016] [Embodiment 1] (Summary of Embodiment 1) Figure 1 is an explanatory diagram illustrating the overview of Embodiment 1. The storage system 1 according to Embodiment 1 is a block storage system. The storage system 1 has one or more drive boxes 6 (#i (i=1, 2, ...N (N is a natural number))). Drive box 6 (#i) is a management unit for drives 62 that houses and manages N drives 62.

[0017] Figure 1(a) shows the data arrangement in storage system 1 before active data is aggregated to the normal drives and inactive data to the power-saving drives for each drive box 6. In Figure 1(a), active and inactive data are distributed across all normal drives. Storage system 1 employs Erasure Coding, for example, a 4D2P RAID (Redundant Arrays of Inexpensive Disks) configuration.

[0018] In Figure 1, normal drives are shown with solid lines and are drives 62 that are powered on or kept powered on. Power-saving drives are shown with dashed lines and are drives 62 that are powered off. Normal drives and power-saving drives are configured for each RAID group.

[0019] In Figure 1, active data is the data stored in drive 62 that is accessed more frequently than a certain level based on I / O requests from the host (not shown), and is indicated by a solid rectangle labeled "1-D0," etc. Inactive data is the data stored in drive 62 that is accessed less frequently than a certain level based on I / O requests from the host, etc., and is indicated by a dashed rectangle labeled "2-D1," etc.

[0020] On the other hand, Figure 1(b) shows the data arrangement in the storage system 1 after active data has been aggregated to the normal drive and inactive data to the power-saving drive for each drive box 6. In Figure 1(b), active data is stored in the normal drive, and inactive data is aggregated and arranged on the power-saving drive. Here, active data that has moved from the power-saving drive to the normal drive is indicated by "3-D0" etc. enclosed in a thick solid rectangle, and inactive data that has moved from the normal drive to the power-saving drive is indicated by "2-P1" etc. enclosed in a thick dashed rectangle.

[0021] (Storage system 1 according to Embodiment 1) Figure 2 shows the configuration of the storage system 1 according to Embodiment 1. The storage system 1 includes a processor 2, a storage device 3, memory 4, a communication interface 5, and one or more drive boxes 6.

[0022] Processor 2 performs overall control of storage system 1, processes I / O requests from host (not shown), and reads and writes data stored in one or more drives 62 housed in each drive box 6.

[0023] The storage device 3 is a non-volatile storage device that stores a drive information table 31, a device information table 32, a block information table 33, and a RAID group information table 34.

[0024] Memory 4 includes an inactive data determination unit 41, a data movement control unit 42, a drive power supply control unit 43, and an I / O processing unit 44. The inactive data determination unit 41, the data movement control unit 42, the drive power supply control unit 43, and the I / O processing unit 44 are functional units that are realized when a program read from a storage area such as a storage device 3 is executed by the processor 2.

[0025] The drive box 6 has a power supply unit 61 and one or more drives 62. The power supply unit 61 supplies power to the drives 62 and turns the power supply to the drives 62 on and off in response to instructions from the processor 2. The drives 62 are non-volatile storage devices such as HDDs (Hard Disk Drives) and SSDs (Solid State Drives).

[0026] (Drive information table 31 according to Embodiment 1) Figure 3 shows the configuration of the drive information table 31 according to Embodiment 1. The drive information table 31 has the following items: drive identification information 311, rated power consumption 312, power supply status 313, normal / power saving 314, capacity 315, maximum access performance 316, active data volume 317, and inactive data volume 318. Furthermore, the drive information table 31 has the following items: cumulative write count 319 and vendor guaranteed write count 320.

[0027] Drive identification information 311 is identification information that identifies drive 62 across drive boxes 6. Rated power consumption 312 indicates the rated power consumption of drive 62 identified by drive identification information 311. Power supply status 313 indicates the on / off state of the power supply, i.e., the power supply, of drive 62 identified by drive identification information 311. Normal / power saving 314 indicates that drive 62 identified by drive identification information 311 is a normal drive if it is "normal", and a power saving drive if it is "power saving". Capacity 315 indicates the total capacity of the storage area of ​​drive 62 identified by drive identification information 311. Maximum access performance 316 indicates the maximum specification throughput of drive 62 identified by drive identification information 311.

[0028] The active data volume 317 indicates the total capacity of active data stored in drive 62 identified by drive identification information 311. The inactive data volume 318 indicates the total capacity of inactive data stored in drive 62 identified by drive identification information 311. The cumulative write count 319 indicates the cumulative number of times data has been written to drive 62 since it started operating. The vendor guaranteed write count 320 is the guaranteed value of the number of writes specified for drive 62 identified by drive identification information 311.

[0029] (Device information table 32 related to Embodiment 1) Figure 4 shows the configuration of the device information table 32 according to Embodiment 1. The device information table 32 stores device information related to the drive box 6. The device information table 32 has device information 321 and item values ​​322. The device information 321 is an item related to the specifications of the drive box 6, such as "number of drives", "backend bandwidth", "RAID type", and "RAID configuration". "Number of drives" is the total number of drives 62 stored in the drive box 6. "Backend bandwidth" is the maximum communication throughput between drives 62 of the drive box 6. "RAID type" indicates the type of RAID adopted in the storage system 1. "RAID configuration" indicates the RAID configuration adopted in the storage system 1.

[0030] (Block information table 33 related to Embodiment 1) Figure 5 shows the configuration of the block information table 33 according to Embodiment 1. The block information table 33 has the following items: block identification information 331, block status 332, last access date and time 333, and drive identification information 334.

[0031] Block identification information 331 is the identification information for each block that is a storage area of ​​drive 62. Block status 332 indicates that the data stored in the block identified by block identification information 331 and drive identification information 334 is active if it is "active" and inactive if it is "inactive". Last access date and time 333 is the date and time when the last access based on an I / O request from the host was performed on the block identified by block identification information 331 and drive identification information 334. Drive identification information 334 is the identification information of the drive (drive identification information 311 in drive information table 31) in which the block identified by block identification information 331 is stored.

[0032] (RAID group information table 34 related to Embodiment 1) Figure 6 shows the configuration of the RAID group information table 34 according to Embodiment 1. The RAID group information table 34 has the following items: RAID group identification information 341, number of components 342, component drive box identification information 343, component drive identification information 344, total capacity 345, total amount of stored data 346, and amount of inactive data 347.

[0033] RAID group identification information 341 is the identification information for the RAID group. Number of components 342 is the number of drives 62 belonging to the RAID group identified by RAID group identification information 341. Components drive box identification information 343 is the identification information for the drive box 6 that houses the drives 62 belonging to the RAID group identified by RAID group identification information 341. Components drive identification information 344 is the identification information for the drives 62 belonging to the RAID group identified by RAID group identification information 341. Total capacity 345 is the total capacity of the storage area of ​​the RAID group identified by RAID group identification information 341. Total amount of stored data 346 is the amount of data stored in the storage area of ​​the RAID group identified by RAID group identification information 341. Amount of inactive data 347 is the total amount of inactive data stored in the storage area of ​​the RAID group identified by RAID group identification information 341.

[0034] In Embodiment 1, for a RAID group configured on a single drive box 6, active data is aggregated to the normal drive, and inactive data is aggregated to the power-saving target drive. Therefore, in Embodiment 1, as shown in rows L1 and L3 of the RAID group information table 34, only RAID groups consisting of a single drive box 6 are processed. In Embodiment 1, as shown in row L2 of the RAID group information table 34, RAID groups consisting of multiple drive boxes 6 are not processed.

[0035] (Drive power off process according to Embodiment 1) Figure 7 is a flowchart showing the drive power-off process according to Embodiment 1. The drive power-off process is performed periodically by the drive power control unit 43 on a per-drive box 6 basis.

[0036] First, in step S11, the drive power control unit 43 refers to the RAID group information table 34 and checks the value of the item 347 for the amount of inactive data in the RAID group within the drive box 6 to be processed.

[0037] Next, in step S12, the drive power control unit 43 determines whether the total amount of inactive data 347 in the RAID group within the drive box 6 to be processed, as confirmed in step S11, exceeds a threshold. If the total amount of inactive data 347 exceeds the threshold (step S12 YES), the drive power control unit 43 moves the process to step S13; otherwise, it returns the process to step S11.

[0038] Next, in step S13, the drive power control unit 43 executes a power saving target drive count calculation process to calculate the number of drives M that are subject to power saving in the drive box 6 to be processed. Details of the power saving target drive count calculation process will be described later with reference to Figure 8.

[0039] Next, in step S14, the drive power control unit 43 executes a drive group determination process to determine the drive group G by selecting the number of drives 62 (M) calculated in step S13 from the drive box 6 to be processed. Details of the drive group determination process will be described later with reference to Figure 9.

[0040] Next, in step S15, the drive power control unit 43 controls the data transfer control unit 42 to move all the data stored in the drive 62 of drive group G to the drives 62 other than those in drive group G in the drive box 6 to be processed.

[0041] Next, in step S16, the drive power control unit 43 removes drive 62 of drive group G from the corresponding RAID group in the RAID group information table 34. Then, in step S17, the drive power control unit 43 creates a new RAID group (low-power RAID group) using drive group G and adds the information to the RAID group information table 34.

[0042] Next, in step S18, the drive power control unit 43 controls the data transfer control unit 42 to move all inactive data in the drive box 6 to the drive 62 of the power-saving RAID group created in step S17.

[0043] Next, in step S19, the drive power control unit 43 turns off the power to all drives 62 of the power-saving RAID group created in step S17. At this time, the drive power control unit 43 changes the power status 313 of the drives 62 whose power has been turned off to "off" and the normal / power saving status 314 to "power saving" in the drive information table 31 (Figure 3).

[0044] (Calculation process for the number of power-saving drives according to Embodiment 1) Figure 8 is a flowchart showing the power saving target drive number calculation process (step S13 (Figure 7)) according to Embodiment 1.

[0045] First, in step S131, the drive power control unit 43 performs a minimum number of drive units calculation process. Details of the minimum number of drive units calculation process will be described later with reference to Figure 9. Next, in step S132, the drive power control unit 43 performs a data storage unit calculation process. Details of the data storage unit calculation process will be described later with reference to Figure 10. Next, in step S133, the drive power control unit 43 performs a performance limit check process. Details of the performance limit check process will be described later with reference to Figure 11.

[0046] Next, in step S134, the drive power control unit 43 compares the minimum number of drives D calculated in step S131 with the number of data storage drives P calculated in step S132, and determines the maximum value of these two values ​​as the number of drives to be targeted for power saving M.

[0047] Next, in step S135, the drive power supply control unit 43 performs a process to calculate the number of unused drives. Details of the process to calculate the number of unused drives will be described later with reference to Figure 12.

[0048] (Calculation process for the minimum number of units according to Embodiment 1) Figure 9 is a flowchart showing the minimum number of units calculation process according to Embodiment 1 (step S131 (Figure 8)).

[0049] First, in step S131a, the drive power control unit 43 calculates a value D (rounded up to the nearest whole number) by dividing the amount of inactive data calculated in step S11 (Figure 7) by the drive capacity (total capacity) of the RAID group.

[0050] Next, in step S131b, the drive power control unit 43 determines whether the value D calculated in step S131a is smaller than the stripe width S of the RAID group. If the value D is smaller than the stripe width S (step S131a YES), the drive power control unit 43 proceeds to step S133c, and if the value D is greater than or equal to the stripe width S (step S131a NO), the minimum number of units calculation process is terminated. In step S133c, the drive power control unit 43 sets D = S.

[0051] (Data storage count calculation process according to Embodiment 1) Figure 10 is a flowchart showing the data storage count calculation process (step S132 (Figure 8)) according to Embodiment 1.

[0052] In step S132a, the drive power control unit 43 refers to the RAID group information table 34 and calculates the ratio R of the total value of the non-active data amount (step S11 (FIG. 7)) to the stored data amount of the RAID group in the drive box 6 to be processed.

[0053] Next, in step S132b, the drive power control unit 43 calculates P = N × R (where N is the total number of drives 62 in the drive box 6 to be processed) and performs rounding up of the fractional part of P (rounding up the decimal part).

[0054] (Performance limit check process according to Embodiment 1) FIG. 11 is a flowchart showing the performance limit check process (step S133 (FIG. 8)) according to Embodiment 1. The performance limit check process determines the number of drives to be power-saving so as to avoid a decrease in write throughput when the number of normal drives is less than a certain number.

[0055] First, in step S133a, the drive power control unit 43 sets the value obtained by dividing the back-end bandwidth of the storage system 1 by the maximum value of the maximum access performance 316 of the drives 62 in the drive box 6 to be processed as Q (rounding up the decimal part).

[0056] Next, in step S133b, the drive power control unit 43 determines whether N - P < Q holds for the total number N of the drives 62 in the drive box 6 to be processed, the P calculated in step S132 (FIG. 8), and the Q calculated in step S133a. When N - P < Q holds (step S133b YES), the drive power control unit 43 transfers the process to step S133c, and when N - P ≥ Q holds (step S133b NO), the performance limit check process ends. In step S133c, the drive power control unit 43 sets P = N - Q.

[0057] (Calculation process of the number of unused drives according to Embodiment 1) FIG. 12 is a flowchart showing the unused drive number calculation process (step S135 (FIG. 8)) according to Embodiment 1. The unused drive number calculation process stops the selection of power-saving target drives and stops the power-saving function when a sufficient number of drives 62 are already unused and powered off.

[0058] First, in step S135a, the drive power control unit 43 calculates the number U of unused drives. That is, for the drive box 6 to be processed, the number U of unused drives (for example, rounded down to the nearest integer) is calculated based on Equation (1). In Equation (1), "a" is the number of drives in the device information table 32, "b" is the value of the item of the total stored data amount 346 in the RAID group information table 34, and "c" is the value of the item of the capacity 315 in the drive information table 31. That is, U is the "number of unused drives 62 in which data that can be calculated based on the total stored data amount is not stored". U = a - (b / c) ··· (1)

[0059] Next, in step S135b, the drive power control unit 43 determines whether M < U holds for the number M of power-saving target drives determined in step S134 (FIG. 8). When M < U holds (step S135b YES), the drive power control unit 43 transfers the process to step S135c. When M ≤ U holds (step S135b NO), the unused drive number calculation process ends. In step S135c, the drive power control unit 43 turns off the power-saving function (the number M of power-saving target drives = 0).

[0060] When the power-saving function is turned off, the following four processes are canceled. That is, the process of determining the power-saving target drives (step S14 (FIG. 7)), the process of moving all the data stored in the power-saving target drives to the normal drives (step S15 (FIG. 7)), the process of moving all the inactive data stored in the normal drives to the power-saving target drives (step S18 (FIG. 7)), and the process of turning off the power supply to the power-saving target drives (step S19 (FIG. 7)).

[0061] (Drive group determination process according to Embodiment 1) Figure 13 is a flowchart of the drive group determination process (step S14 (Figure 7)) according to Embodiment 1.

[0062] First, in step S141, the drive power control unit 43 performs a candidate drive extraction process. Details of the candidate drive extraction process will be described later with reference to Figure 14. Next, in step S142, the drive power control unit 43 performs a drive group selection process. Details of the drive group selection process will be described later with reference to Figure 15.

[0063] (Candidate drive extraction process according to Embodiment 1) Figure 14 is a flowchart of the candidate drive extraction process (step S141 (Figure 13)) according to Embodiment 1. The candidate drive extraction process aims to equalize the lifespan of drives by designating drive 62, which has a low remaining write cycle, as a power-saving target drive that stores inactive data and is shut down, thereby reducing maintenance costs by enabling drive replacement at the same time.

[0064] First, in step S141a, the drive power control unit 43 extracts M drives 62 (drive identification information 311) for the drive box 6 to be processed, in descending order of the cumulative write count value T of the cumulative write count 319 (drive information table 31 (Figure 3)).

[0065] Next, in step S141b, the drive power control unit 43 extracts a drive 62 (drive identification information 311) that meets a predetermined condition, separately from the M drives 62. The predetermined condition is that the difference between the minimum cumulative write count T of the M drives 62 extracted in step S141a and the current cumulative write count T is less than a predetermined percentage (e.g., 1%) of the vendor-guaranteed write count W for the vendor-guaranteed write count 320 (Figure 3). Here, let's assume that the G drives 62 have been extracted.

[0066] Next, in step S141c, the drive power control unit 43 selects the M-unit drive 62 extracted in step S141a and the G-unit drive 62 extracted in step S141b as candidate drives.

[0067] (Drive selection process according to Embodiment 1) Figure 15 is a flowchart of the drive selection process (step S142 (Figure 13)) according to Embodiment 1. The drive selection process aims to reduce the processing time for data transfer when moving data stored from a power-saving drive to a normal drive, and to reduce the number of write cycles to the drive by selecting a drive with the smallest amount of stored data.

[0068] First, in step S142a, the drive power control unit 43 sorts the M units + G units of drives 62 extracted in step S141 (Figure 13) in ascending order of active data amount + inactive data amount based on the drive information table 31.

[0069] Next, in step S142b, the drive power control unit 43 selects M drives 62 from the drives 62 sorted in step S142a in ascending order of active data amount + inactive data amount. These M drives 62 are designated as drive group G.

[0070] (Inactive data access processing when the drive power is turned off, according to Embodiment 1) Figure 16 is a flowchart illustrating the inactive data access process when the drive power is turned off according to Embodiment 1. The inactive data access process when the drive power is turned off is executed when a host (not shown) makes an access request based on an I / O request or the like for inactive data stored in drive 62, which has been powered off by the drive power-off process (Figure 7).

[0071] First, in step S21, the I / O processing unit 44 confirms the block identification information of the data to be accessed. Next, in step S22, the I / O processing unit 44 determines whether the data whose block identification information was confirmed in step S21 is inactive data. If the data is inactive (step S22 YES), the I / O processing unit 44 moves the process to step S23, and if the data is active (step S22 NO), it moves the process to step S29.

[0072] In step S23, the I / O processing unit 44 searches for drive O, the location where the inactive data to be accessed is stored. Next in step S24, the I / O processing unit 44 controls the drive power control unit 43 to turn on the power to drive O, which was found in step S23. Next in step S25, the I / O processing unit 44 waits until drive O becomes accessible.

[0073] Next, in step S26, the I / O processing unit 44 determines whether drive O is accessible. The I / O processing unit 44 refers to the power status 313 in the drive information table 31 and determines whether drive O is accessible (power on). If drive O is accessible (step S26YES), the I / O processing unit 44 moves to step S27, and if drive O is not accessible (step S26NO), it returns to step S25.

[0074] In step S27, the I / O processing unit 44 accesses the inactive data stored in drive O. Next, in step S28, the I / O processing unit 44 controls the drive power control unit 43 to turn off the power to drive O after access to drive O is complete.

[0075] Meanwhile, in step S29, the I / O processing unit 44 performs normal access to the data that was determined to be active data in step S22.

[0076] (Inactive data addition and storage process according to Embodiment 1) Figure 17 is a flowchart showing the inactive data addition and storage process according to Embodiment 1. The inactive data addition and storage process is executed at predetermined intervals after the power to the drive 62 is turned off by the drive power off process (Figure 7).

[0077] First, in step S31, the inactive data determination unit 41 executes an inactive data determination process. Details of the inactive data determination process will be described later with reference to Figure 18. Next, in step S32, the inactive data determination unit 41 determines whether the increase amount Δ of the data amount determined to be inactive in step S31 compared to the data amount at the time of the previous determination exceeds a threshold. If the increase amount Δ exceeds the threshold (step S32 YES), the inactive data determination unit 41 moves the process to step S33, and if the increase amount Δ does not exceed the threshold (step S32 NO), it returns the process to step S31.

[0078] In step S33, the inactive data determination unit 41 refers to the drive information table 31 (Figure 3) and turns on the power to all power-saving target drives that are currently powered off. Next, in step S34, the inactive data determination unit 41 waits until the power-saving target drives that were powered on in step S33 become accessible.

[0079] Next, in step S35, the inactive data determination unit 41 refers to the power status 313 of the drive information table 31 and determines whether the power-saving target drive is accessible (powered on). If the power-saving target drive is accessible (step S35YES), the inactive data determination unit 41 moves the process to step S36, and if the power-saving target drive is not accessible (step S35YES), it returns the process to step S34.

[0080] In step S36, the inactive data determination unit 41 controls the data movement control unit 42 to move the data newly determined to be inactive in step S31 to the power-saving target drive. Next, in step S37, the inactive data determination unit 41 controls the data movement control unit 42 to move the data newly determined to be active in step S31 to the normal drive. Next, in step S38, the inactive data determination unit 41 controls the drive power supply control unit 43 to turn off the power to the power-saving target drive.

[0081] (Inactive data determination process according to Embodiment 1) Figure 18 is a flowchart of the inactive data determination process (step S31 (Figure 7)) according to Embodiment 1. The inactive data determination process is performed for all data in which the last access date and time 333 is recorded in the block information table 33.

[0082] First, in step S31a, the inactive data determination unit 41 refers to the block information table 33 and confirms the last access date and time 333 of each block (block identification information 331). Next, in step S31b, the inactive data determination unit 41 determines whether the last access date and time 333 has been more than a predetermined period (for example, one month) since the current date and time (whether the access frequency is above or below a certain level). If the last access date and time 333 has been more than a predetermined period since the current date and time (step S31b YES), the inactive data determination unit 41 moves to step S31c, and if less than a predetermined period has been elapsed (step S31b NO), it moves to step S31d.

[0083] In step S31c, the inactive data determination unit 41 determines in step S31b that data whose last access date and time 333 is more than a predetermined period from the current date and time is inactive data, and updates the corresponding record in the block information table 33. On the other hand, in step S31d, the inactive data determination unit 41 determines in step S31b that data whose last access date and time 333 is less than a predetermined period from the current date and time is active data, and updates the corresponding record in the block information table 33.

[0084] (Modification 1 of Embodiment 1) In Embodiment 1, inactive data is aggregated to a power-saving target drive in units of 62 drives within a single RAID group. However, Embodiment 1 can also be applied when aggregating inactive data to a certain RAID group on a RAID group basis within the storage system 1, and the amount of inactive data exceeds the capacity of the target RAID group.

[0085] In other words, if the total amount of inactive data 347 (Figure 6) stored in the drives 62 that make up multiple RAID groups of storage system 1 exceeds the total capacity 345 (Figure 6) of the target RAID group, the following process is performed: A new RAID group is created with a number of drives 62 that can store all the inactive data stored in storage system 1, and the drives 62 that make up this RAID group are designated as power-saving drives. Then, all active data is aggregated to the normal drives, and all inactive data is aggregated to the power-saving drives, and the power to the power-saving drives is turned off. In this way, all inactive data across multiple RAID groups can be aggregated to the power-saving drives, thereby reducing the power consumption of the power-saving drives.

[0086] (Modification 2 of Embodiment 1) In Embodiment 1, multiple drives 62 housed in the drive box 6 are targeted for processing, and inactive data is aggregated to the drive targeted for power saving. However, the system is not limited to this; multiple drives 62 may be targeted for processing and inactive data aggregated to the drive targeted for power saving, regardless of whether they are housed in the drive box 6 or whether they constitute a RAID group.

[0087] (Effects of Embodiment 1) In Embodiment 1, the normal drive and the power-saving target drive are determined based on the amount of inactive data stored on multiple drives. Then, all data stored on the power-saving target drive is moved to the normal drive, and then all inactive data stored on the normal drive is moved to the power-saving target drive. After that, the power supply to the power-saving target drive is turned off. Therefore, even if it is not possible to consolidate all active and inactive data onto the power-saving target drive, power saving effects are efficiently achieved by consolidating inactive data, which has a low probability of being accessed, onto the power-saving target drive and then turning off the power to the power-saving target drive.

[0088] Furthermore, in Embodiment 1, a drive to be used for power saving is selected from multiple drives based on the drive's write performance. This prevents problems such as a decrease in write throughput due to a small number of drives in the RAID group of the drive to be used for power saving.

[0089] Furthermore, in Embodiment 1, power-saving target drives are selected from multiple drives based on the lifespan of each drive, which is determined by the actual number of writes to the drive and the number of writeables specified in the product specifications. Therefore, by using drives with fewer remaining writes to store inactive data that is accessed infrequently, the remaining lifespan of each drive is equalized, and maintenance costs can be reduced by ensuring that the replacement timing due to the lifespan of each drive is the same.

[0090] In Embodiment 1, the drive to be used for power saving is selected from multiple drives based on the order in which the amount of inactive data stored on each drive is smallest. When the power saving function is enabled, it is necessary to move the stored data from the drive to be used for power saving to the normal drive. By selecting a drive with a small amount of stored data as the drive to be used for power saving, the time required for moving stored data can be reduced, and the number of writes to the drive can be reduced, thereby extending the drive's lifespan.

[0091] In Embodiment 1, when an I / O request is received for inactive data stored in a power-saving drive whose power supply from the host has been turned off, the power supply to the power-saving drive is turned on. Then, after the power supply to the power-saving drive is turned on, the inactive data is accessed, and when access to the inactive data is finished, the power supply to the power-saving drive is turned off. Therefore, even when an I / O request is received for inactive data stored in a power-saving drive, I / O processing can be performed.

[0092] In Embodiment 1, it is determined at predetermined intervals whether each piece of data stored on multiple drives is active or inactive. It is then determined whether the increase in the amount of inactive data determined this time exceeds a threshold compared to the amount of inactive data determined last time. If the increase exceeds the threshold, the power supply to the drive targeted for power saving is turned off, and after turning the power supply to the drive targeted for power saving is turned on, the inactive data corresponding to the increase is moved from the normal drive to the drive targeted for power saving. After moving the inactive data corresponding to the increase from the normal drive to the drive targeted for power saving, the power supply to the drive targeted for power saving is turned off. Therefore, even if inactive data increases over time, it can be moved to the drive targeted for power saving as needed, and the capacity of the normal drive can be used efficiently.

[0093] In Embodiment 1, the number of drives targeted for power saving is compared with the number of unused drives that do not contain data and are powered off. If the number of drives is less than the number of unused drives, the process of determining which drives are targeted for power saving and the process of moving all data stored on the drives targeted for power saving to the normal drives are canceled. Furthermore, the process of moving all inactive data stored on the normal drives to the drives targeted for power saving and the process of turning off the power supply to the drives targeted for power saving are also canceled. Thus, the power saving function can be activated under conditions where it is effective, and the power saving function can be disabled under conditions where it is not effective, thereby eliminating unnecessary processing.

[0094] [Embodiment 2] In Embodiment 1, for a RAID group configured in a single drive box 6, active data is aggregated to the normal drive and inactive data to the power-saving target drive for each drive box 6, and the power to the power-saving target drive is turned off. In contrast, in Embodiment 2, for a RAID group spanning multiple drive boxes 6, active data is aggregated to the normal drive and inactive data to the power-saving target drive across the drive boxes 6. The power to the power-saving target drive is then turned off. The following description of Embodiment 2 will focus on the differences from Embodiment 1.

[0095] In Embodiment 2, the target of processing is a RAID group consisting of multiple drive boxes 6, as shown in row L2 of the RAID group information table 34 (Figure 6).

[0096] (Summary of Embodiment 2) Figure 19 is an explanatory diagram illustrating the overview of Embodiment 2. In Figure 19, a normal drive is shown by a solid line and is a drive 62 that is powered on or kept powered on. A power-saving target drive is shown by a dashed line and is a drive 62 that is powered off.

[0097] Figure 19(a) shows the data arrangement in storage system 1B before active data was aggregated to normal drives and inactive data to power-saving drives for the RAID group configured in drive box 6 (#1, #2). In Figure 19(a), active and inactive data are distributed across all normal drives in drive box 6 (#1, #2).

[0098] On the other hand, Figure 19(b) shows the data arrangement in storage system 1B after active data has been aggregated to normal drives and inactive data to power-saving drives across drive boxes 6 (#1, #2). In Figure 19(b), active data is stored in the normal drives (drives 62(#1~#22)) of drive 62 in drive box 6 (#1). Inactive data is aggregated in the power-saving drives (drives 62(#23~#24)) of drive box 6 (#1) and the power-saving drives (drives 62(#25~#38)) of drive box 6 (#2).

[0099] (Storage system 1B according to embodiment 2) Figure 20 shows the configuration of storage system 1B according to Embodiment 2. Compared to storage system 1 according to Embodiment 1, storage system 1B further includes a power supply control unit 45 in the memory 4. The power supply control unit 45 controls the on / off switching of the power supply function unit 61 provided in the drive box 6.

[0100] (Drive group determination process according to Embodiment 2) Figure 21 is a flowchart showing the drive group determination process according to Embodiment 2. The drive group determination process according to Embodiment 2 is executed in place of the drive group determination process according to Embodiment 1 (Figure 13).

[0101] First, in step S140B, the drive power supply control unit 43 performs a power-saving drive box and drive determination process. Details of the power-saving drive box and drive determination process will be described later with reference to Figure 22.

[0102] Next, in step S141B, the drive power control unit 43 performs a candidate drive extraction process. Details of the candidate drive extraction process will be described later with reference to Figure 23. Next, in step S142B, the drive power control unit 43 performs a drive selection process. Details of the drive selection process will be described later with reference to Figure 24.

[0103] (Power saving drive box and drive housing determination process according to Embodiment 2) Figure 22 is a flowchart showing the power-saving drive box and drive determination process (step S140B (Figure 21)) according to Embodiment 2. In the power-saving drive box and drive determination process, for drive boxes 6 having a number of drives 62 less than or equal to the number M calculated in the power-saving target drive number calculation process (step S13 (Figure 7)), all drives 62 are designated as power-saving target drives and their power supply is turned off. Furthermore, power supply to the power supply function unit 61 of the drive box 6 is turned off to achieve further power savings.

[0104] First, in step S140Ba, the drive power control unit 43 determines whether there is a drive box 6 that can accommodate M or fewer drives 62. If there is a drive box 6 that can accommodate M or fewer drives 62 (step S140BaYES), the drive power control unit 43 proceeds to step S140Bb. On the other hand, if there is no drive box 6 that can accommodate M or fewer drives 62 (step S140BaNO), the drive power control unit 43 proceeds to step S140Bc.

[0105] In step S140Bb, the drive power control unit 43 selects b drives 62 to be housed in the drive box 6 which houses M or fewer drives 62 in step S140Ba. On the other hand, in step S140Bc, the drive power control unit 43 sets b=0.

[0106] (Candidate drive extraction process according to Embodiment 2) Figure 23 is a flowchart of the candidate drive extraction process (step S141B (Figure 21)) according to Embodiment 2.

[0107] First, in step S141Ba, the drive power control unit 43 extracts Mb drives 62 (drive identification information 311) from the drive information table 31 for the drive box 6 to be processed, in descending order of the cumulative write count value T for the cumulative write count 319. The drive box 6 to be processed here is the drive box 6 that was determined in step S140Ba (Figure 22) to have more than M drives 62 to house.

[0108] Next, in step S141Bb, the drive power control unit 43 extracts drives 62 (drive identification information 311) that meet predetermined conditions, separately from the Mb-class drives 62. The predetermined conditions are that the difference between the minimum cumulative write count T of the Mb-class drives 62 extracted in step S141Ba and the current cumulative write count T is less than a predetermined percentage (e.g., 1%) of the vendor-guaranteed write count W for the vendor-guaranteed write count 320. Here, let's assume that G-class drives 62 have been extracted.

[0109] Next, in step S141Bc, the drive power control unit 43 selects the Mb-class drive 62 extracted in step S141Ba and the G-class drive 62 extracted in step S141Bb as candidate drives.

[0110] (Drive selection process according to Embodiment 2) Figure 24 is a flowchart showing the drive selection process (step S142B (Figure 21)) according to Embodiment 2.

[0111] First, in step S142Ba, the drive power control unit 43 sorts the Mb+G drives 62 extracted in step S141B (Figure 21) in ascending order of active data amount + inactive data amount based on the drive information table 31.

[0112] Next, in step S142Bb, the drive power control unit 43 selects drives 62 in the Mb range from the drives 62 sorted in step S142Ba in ascending order of active data amount + inactive data amount.

[0113] Furthermore, the Mb drive 62 selected in step S142Bb and the b drive 62 selected in step S140Bb (Figure 22) are powered off in step S19 of the drive power-off process (Figure 7). In addition, the b drive 62 selected in step S140Bb is powered off in step S19 of the drive power-off process, and the power supply function unit 61 of the drive box 6 that houses the b drive 62 is also powered off.

[0114] (Effects of Embodiment 2) In Embodiment 2, after turning off the power supply to the drive targeted for power saving, the power supply to the drive box containing all of the drives targeted for power saving is also turned off. Therefore, since the power to the drive box can be turned off in addition to the drives, an even greater power saving effect can be achieved.

[0115] [Embodiment 3] In Embodiment 1, active data is aggregated to the normal drive, and inactive data is aggregated to the power-saving target drive, and the power to the power-saving target drive is turned off. In contrast, in Embodiment 3, active data + inactive data are aggregated to the normal drive, and the power to the power-saving target drive is turned off. The following description of Embodiment 3 will focus on the differences from Embodiment 1.

[0116] (Drive power off process according to Embodiment 3) Figure 25 is a flowchart showing the drive power-off process according to Embodiment 3. The drive power-off process according to Embodiment 3 is performed in place of the drive power-off process according to Embodiment 1 (Figure 7). The drive power-off process according to Embodiment 3 is performed periodically by the drive power control unit 43 in units of drive boxes 6.

[0117] First, in step S11C, the drive power control unit 43 refers to the RAID group information table 34 and calculates the total value r11 of the total amount of data 346 stored in the RAID group within the drive box 6 to be processed.

[0118] Next, in step S12C, the drive power control unit 43 determines whether r12 = (sum of all capacities of the RAID groups in the drive box 6 to be processed - r11) is equal to or greater than the capacity of one drive 62. If r12 is equal to or greater than the capacity of one drive 62 (step S12CYES), the drive power control unit 43 moves the process to step S13C, and if r12 is less than the capacity of one drive 62 (step S12CNO), the process returns to step S11C.

[0119] Next, in step S13C, the drive power control unit 43 calculates the number of power-saving drives m in the drive box 6 to be processed. In step S13C, the drive power control unit 43 refers to the drive information table 31 and divides the r12 calculated in step S12C by the capacity value of each drive 62, which is 315 (rounded down to the nearest whole number), and uses this value as the number of power-saving drives m.

[0120] Next, in step S14C, the drive power control unit 43 executes a drive group determination process to determine the drive group G by selecting the number of drives 62 (m) calculated in step S13C from the drive box 6 to be processed. This drive group determination process is the same as the drive group selection process according to Embodiment 1 (Figure 13), but with the number of drives 62 "M" replaced by "m".

[0121] Next, in step S15C, the drive power control unit 43 controls the data transfer control unit 42 to move all the data stored in the drive 62 of drive group G to the drives 62 other than those in drive group G in the drive box 6 to be processed.

[0122] Next, in step S19C, the drive power control unit 43 turns off the power to all drives 62 of the drive group G determined in step S14C.

[0123] (Effects of Embodiment 3) In Embodiment 3, a normal drive and a drive to be power-saving are determined on a drive-by-drive basis, data is moved from the drive to be power-saving to the normal drive, and the power supply to the drive to be power-saving is turned off. As a result, unused areas distributed across each drive are consolidated into one or more drives, and the power is turned off on a drive-by-drive basis, thus enabling fine-grained power saving effects corresponding to the size of the unused area.

[0124] [Embodiment 4] In Embodiment 3, active data and inactive data are aggregated into a normal drive in a more granular manner, using 62 drives per RAID group, and the power to the power-saving target drive is turned off. In contrast, in Embodiment 4, active data and inactive data are aggregated into a normal RAID group using 62 drives across multiple pools (RAID groups). Then, the power to the power-saving target drive 62 that constitutes the pool (RAID group) that does not store data is turned off.

[0125] In other words, the unused space of the 62 drives distributed across each pool (RAID group) is consolidated into a single pool (RAID group), and the power to the power-saving drives that make up this pool is turned off.

[0126] Here, a normal RAID group is a RAID group in which the drives 62 constituting the RAID group are powered on or kept powered on. A power-saving RAID group is a RAID group in which the drives 62 constituting the RAID group are powered off. The following description of Embodiment 4 will focus on the differences from Embodiment 3.

[0127] (Drive power off process according to Embodiment 4) Figure 26 is a flowchart showing the drive power-off process according to Embodiment 4. The drive power-off process according to Embodiment 4 is performed in place of the drive power-off process according to Embodiment 3 (Figure 25). The drive power-off process according to Embodiment 4 is performed periodically by the drive power control unit 43 for multiple RAID groups to be processed.

[0128] First, in step S11D, the drive power control unit 43 refers to the RAID group information table 34 and calculates the total amount of stored data r21 for each of the multiple RAID groups to be processed.

[0129] Next, in step S12D, the drive power control unit 43 determines whether R22 = (sum of the capacities of multiple RAID groups - r21) is equal to or greater than the capacity of one drive 62. If r22 is equal to or greater than the capacity of one drive 62 (step S12DYES), the drive power control unit 43 moves the process to step S13D, and if r22 is less than the capacity of one drive 62 (step S12DNO), the process returns to step S11D.

[0130] Next, in step S13D, the drive power control unit 43 calculates the number of power-saving target RAID groups q in the drive box 6 to be processed. In step S13D, the drive power control unit 43 refers to the RAID group information table 34 and takes the value obtained by dividing r22 calculated in step S12D by the sum of the total capacities 345 of the multiple RAID groups (rounded down to the nearest whole number) as the number of power-saving target RAID groups q.

[0131] Next, in step S14D, the drive power control unit 43 executes a RAID group determination process to select the number of q RAID groups calculated in step S13D and determine the RAID group group RG. In this RAID group determination process, the RAID group group RG is determined in ascending order of the total amount of stored data 346.

[0132] Next, in step S15D, the drive power control unit 43 controls the data transfer control unit 42 to move all the data stored in the drive 62 of RAID group RG to the drive 62 of RAID groups other than RAID group RG.

[0133] Next, in step S19D, the drive power control unit 43 designates the RAID group group RG determined in step S14D as the RAID group to be power-saving, and turns off the power to all drives 62 that make up the RAID group to be power-saving.

[0134] (Effects of Embodiment 4) In Embodiment 4, normal drives and power-saving target drives are determined on a pool (RAID group) basis, data is moved from the power-saving target drives to the normal drives, and power supply to the power-saving target drives is turned off. Therefore, unused areas distributed across each pool are consolidated into one pool, and the drives are powered off on a pool-by-pool basis, resulting in significant power savings.

[0135] Although several embodiments have been described above, these are merely illustrative examples for explaining the present invention and are not intended to limit the scope of the present invention to these embodiments only. The present invention can also be implemented in various other forms, such as forms in which some of the components of the above embodiments are omitted, forms in which at least some of the components are replaced, forms in which components are added, or forms that combine some or all of the embodiments. [Explanation of Symbols]

[0136] 1B: Storage system, 2: Processor, 4: Memory, 6: Drive box, 41: Inactive data detection unit, 42: Data movement control unit, 43: Drive power control unit, 44: I / O processing unit, 45: Power supply function unit power control unit, 61: Power supply function unit, 62: Drive

Claims

1. A storage system having multiple drives and processors that are accessed in response to I / O requests from a host, The aforementioned processor, Based on the amount of data stored in the aforementioned multiple drives, the system determines which of the multiple drives will have its power supply turned on (normal drives) and which will have its power supply turned off (power-saving drives). All data stored in the aforementioned power-saving drive is moved to the aforementioned normal drive. After moving all data stored on the power-saving drive to the normal drive, the power supply to the power-saving drive is turned off. A storage system characterized by the following features.

2. A storage system according to claim 1, The aforementioned processor, Based on the last access date and time for accessing each piece of data stored on the aforementioned multiple drives, the system manages the determination results that determine whether each piece of data is active data accessed at a certain frequency or more, or inactive data accessed at a certain frequency or less. Based on the amount of data of the inactive data stored in the plurality of drives, the normal drives and the power-saving target drives are determined. After moving all data stored on the power-saving drive to the normal drive, all inactive data stored on the normal drive is moved to the power-saving drive. After moving all the inactive data stored in the normal drive to the power-saving drive, the power supply to the power-saving drive is turned off. A storage system characterized by the following features.

3. A storage system according to claim 2, The aforementioned processor, Based on the write performance of the aforementioned drive, the power-saving target drive is selected from the plurality of drives. A storage system characterized by the following features.

4. A storage system according to claim 2, The aforementioned processor, Based on the lifespan of each drive, which is determined by the actual number of writes and the number of writes specified in the product specifications for each drive, the power-saving target drive is selected from the plurality of drives. A storage system characterized by the following features.

5. A storage system according to claim 2, The aforementioned processor, The power-saving target drive is selected from the plurality of drives based on the order in which the amount of inactive data stored in each drive is smallest. A storage system characterized by the following features.

6. A storage system according to claim 2, The aforementioned processor, When an I / O request is made to the inactive data stored in the power-saving drive, which has its power supply turned off by the host, the power supply to the power-saving drive is turned on. After the power supply to the power-saving target drive is turned on, access is made to the inactive data. When access to the inactive data ends, the power supply to the power-saving drive is turned off. A storage system characterized by the following features.

7. A storage system according to claim 2, The aforementioned processor, At predetermined intervals, it is determined whether each piece of data stored in the plurality of drives is active data or inactive data. Determine whether the increase in the inactive data determined this time compared to the previously determined inactive data exceeds the threshold. If the aforementioned increase exceeds the threshold, the power supply to the power-saving target drive is turned on. After turning on the power supply to the power-saving target drive, the inactive data corresponding to the increase is moved from the normal drive to the power-saving target drive. After moving the inactive data corresponding to the aforementioned increase amount from the normal drive to the power-saving target drive, the power supply to the power-saving target drive is turned off. A storage system characterized by the following features.

8. A storage system according to claim 2, The aforementioned processor, The number of drives targeted for power saving is compared with the number of unused drives that do not contain any data. If the number of drives is less than the number of drives, the process of determining the drive to be power-saving, the process of moving all data stored on the drive to be power-saving to the normal drive, the process of moving all inactive data stored on the normal drive to the drive to be power-saving, and the process of turning off the power supply to the drive to be power-saving are canceled. A storage system characterized by the following features.

9. A storage system according to claim 2, The aforementioned plurality of drives are housed in a plurality of drive boxes, each containing a predetermined number of such drives. The aforementioned processor, After turning off the power supply to the aforementioned power-saving target drive, the power supply to the drive box in which all of the drives housed among the plurality of drive boxes are the power-saving target drives is turned off. A storage system characterized by the following features.

10. A storage system according to claim 1, The aforementioned processor, Based on the amount of data stored in each of the aforementioned multiple drives, the normal drives and the power-saving target drives are determined on a per-drive basis. All data stored on the aforementioned power-saving drive is moved to the standard drive. After moving all the data stored on the aforementioned power-saving drive to the normal drive, the power supply to the power-saving drive is turned off. A storage system characterized by the following features.

11. A storage system according to claim 1, The aforementioned processor, Based on the amount of data stored in each pool on the aforementioned multiple drives, the normal drives and the power-saving drives are determined on a pool-by-pool basis. All data for each pool stored in the power-saving drive is moved to the normal drive. After moving all the data for each pool stored on the power-saving drive to the normal drive, the power supply to the power-saving drive is turned off. A storage system characterized by the following features.

12. A power control method for a storage system having multiple drives and processors that are accessed in response to I / O requests from a host, The aforementioned processor, Based on the amount of data stored in the aforementioned multiple drives, the system determines which of the multiple drives will have its power supply turned on (normal drives) and which will have its power supply turned off (power-saving drives). All data stored in the aforementioned power-saving drive is moved to the aforementioned normal drive. After moving all data stored on the power-saving drive to the normal drive, the power supply to the power-saving drive is turned off. A power control method in a storage system, characterized by having each of the following processes.

Images