STORAGE SYSTEM

• US 20110264854A1
• Filed: 03/24/2009
• Published: 10/27/2011
• Est. Priority Date: 03/24/2009
• Status: Active Grant

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Ouchi, Katsumi
• Granted Patent: US 8,356,139 B2
• Time in Patent Office: Days
• US Class Current: 711/114
• CPC Class Codes: G06F 1/3268 Power saving in hard disk drive
• CPC Class Codes: G06F 11/3423 where the assessed time is active or idle time
• CPC Class Codes: G06F 11/3485 for I/O devices
• CPC Class Codes: G06F 3/0616 in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]
• CPC Class Codes: G06F 3/0625 Power saving in storage systems
• CPC Class Codes: G06F 3/0634 by changing the state or mode of one or more devices
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

To maintain reliability even when the repetition of the power saving control for storage systems occurs frequently due to the access from the host computer. During the course of controlling the operation mode of the drives, the controller measures the start/stop cycle count and the load/unload cycle count, calculates the S/S wear-out rate showing the ratio of the measured value of the start/stop cycle count to the first upper limit value and the L/U wear-out rate showing the ratio of the measured value of the load/unload cycle count to the second upper limit value, and for the drives in the idle operation mode, selects the low-speed rotation operation mode if the S/S wear-out rate is greater than the L/U wear-out rate, or selects the standby operation mode if the S/S wear-out rate is less than the L/U wear-out rate.

Claims

1. A storage system, comprising:

a plurality of storages for storing data; and

a controller connected to a host computer via a network and which identifies access from the host computer, specifies a storage to be subject to the access among the plurality of storages, sends and receives data to and from the specified storage, and controls an operation mode of the plurality of storages;

wherein the controller controls the operation mode of either an idle operation mode which causes the storage to enter an idle status, or a plurality of power saving operation modes which cause the storages to enter a power saving status that saves more power than the idle operation mode;

wherein the controller includes a local memory for storing the respective upper limit values of a plurality of control cycle counts showing an execution count of a plurality of control cycles in which the idle operation mode and each of the power saving operation modes make a circuit; and

wherein, during the course of controlling the operation mode, the controller measures each of the control cycle counts, and selects a power saving operation mode among the plurality of power saving operation modes based on the measured value for deconcentrating the increase in the measured value of each of the control cycle counts in relation to each of the upper limit values.

2. The storage system according to claim 1,

wherein the plurality of power saving operation modes include a low-speed rotation operation mode for causing the storages to enter a rotation status that is slower than the idle operation mode, and a standby operation mode for causing the storages to enter a rotation stop status;

wherein the local memory stores as the plurality of upper limit values;

a first upper limit value as an upper limit value of a first power saving control cycle count showing an execution count of a first power saving control cycle (start/stop cycle) in which the idle operation mode and the standby operation mode make a circuit; and

a second upper limit value as an upper limit value of a second power saving control cycle count showing an execution count of a second power saving control cycle (load/unload cycle) in which the idle operation mode and the power saving mode of either the low-speed rotation operation mode or the standby operation mode make a circuit;

wherein, during the course of controlling the operation mode of the storages, the controller measures the first power saving control cycle count and the second power saving control cycle count, and calculates a first wear-out rate (S/S wear-out rate) showing the ratio of a measured value of the first power saving control cycle count and the first upper limit value stored in the local memory, and a second wear-out rate (L/U wear-out rate) showing the ratio of a measured value of the second power saving control cycle count and the second upper limit value stored in the local memory;

compares the calculated the first wear-out rate (S/S wear-out rate) and the second wear-out rate (UU wear-out rate); and

for storages in the idle operation mode;

selects the low-speed rotation operation mode if the first wear-out rate (S/S wear-out rate) is greater than the second wear-out rate (UU wear-out rate); and

selects the standby operation mode if the first wear-out rate (S/S wear-out rate) is less than the second wear-out rate (L/U wear-out rate).

3. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG), and the plurality of RAID groups (RG) have a plurality of logical devices (LDEV) as virtual logical areas configured therein;

wherein, during the course of controlling an operation mode of storages belonging to each of the RAID groups (RG), the controller measures the access interval to each of the logical devices (LDEV) belonging to each of the RAID groups (RG), and measures the frequency in which the measured access interval exceeds a prescribed setting time or the cumulative count in which the measured access interval exceeds the setting time, sets the logical device (LDEV) with the maximum frequency or cumulative count as a relocation target logical device (LDEV), selects a relocation destination RAID group among RAID groups (RG) other than the RAID group to which the relocation target logical device (LDEV) belongs, sets one of the logical devices (LDEV) belonging to the relocation destination RAID group as a replacement target logical device (LDEV), relocates the relocation target logical device (LDEV) to the relocation destination RAID group, and relocates the replacement target logical device (LDEV) to the RAID group (RG) to which the relocation target logical device (LDEV) used to belong.

4. The storage system according to claim 1,

wherein the plurality of power saving operation modes include a low-speed rotation operation mode for causing the storages to enter a rotation status that is slower than the idle operation mode, and a standby operation mode for causing the storages to enter a rotation stop status;

wherein the local memory stores, as the plurality of upper limit values, a first upper limit value as an upper limit value of a first power saving control cycle count showing an execution count of a first power saving control cycle (start/stop cycle) in which the idle operation mode and the standby operation mode make a circuit, and a second upper limit value as an upper limit value of a second power saving control cycle count showing an execution count of a second power saving control cycle (load/unload cycle) in which the idle operation mode and the power saving mode of either the low-speed rotation operation mode or the standby operation mode make a circuit;

wherein, during the course of controlling the operation mode of the storages belonging to each of the RAID groups (RG), the controller measures the first power saving control cycle count and the second power saving control cycle count;

compares a measured value of the first power saving control cycle count and a measured value of the second power saving control cycle count; and

for storages in the idle operation mode;

selects the low-speed rotation operation mode if the increase rate of the measured value of the first power saving control cycle count in relation to the first upper limit value is greater than the increase rate of the measured value of the second power saving control cycle count in relation to the second upper limit value; and

selects the standby operation mode if the increase rate of the measured value of the first power saving control cycle count in relation to the first upper limit value is less than the increase rate of the measured value of the second power saving control cycle count in relation to the second upper limit value.

5. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG);

wherein the plurality of power saving operation modes include a low-speed rotation operation mode for causing the storages to enter a rotation status that is slower than the idle operation mode, and a standby operation mode for causing the storages to enter a rotation stop status;

wherein, during the course of controlling the operation mode of the storages belonging to each of the RAID groups (RG), the controller measures the access interval for causing the storages belonging to each of the RAID groups (RG) to enter an active operation mode, measures the count in which the measured access interval will become longer than the power saving control waiting time for causing the storages belonging to each of the RAID groups (RG) to enter the low-speed operation mode or the standby operation mode, and determines whether the frequency of the measured count is greater than a setting value; and

for the storages of the RAID groups (RG) in the idle operation mode;

selects the low-speed rotation operation mode if the frequency of the measured count is greater than the setting value; and

selects the standby operation mode if the frequency of the measured count is less than the setting value.

6. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups;

wherein the plurality of power saving operation modes include at least a standby operation mode for causing the storages to enter a rotation stop status;

wherein, during the course of controlling the operation mode of the storages belonging to each of the RAID groups (RG), the controller measures the access interval for causing the storages belonging to each of the RAID groups (RG) to enter an active operation mode, measures the count in which the measured access interval will become longer than the power saving control waiting time for causing the storages belonging to each of the RAID groups (RG) to enter the standby operation mode, and determines whether the frequency of the measured count is greater than a setting value; and

for the storages of the RAID groups (RG) in the idle operation mode;

maintains the idle operation mode if the frequency of the measured count is greater than the setting value; and

selects the standby operation mode if the frequency of the measured count is less than the setting value.

7. The storage system according to claim 1,

wherein the plurality of storage are divided into a plurality of RAID groups;

wherein the plurality of power saving operation modes include a low-speed rotation operation mode for causing the storages to enter a rotation status that is slower than the idle operation mode, and a standby operation mode for causing the storages to enter a rotation stop status;

wherein the local memory stores, as the plurality of upper limit values, a first upper limit value as an upper limit value of a first power saving control cycle count showing an execution count of a first power saving control cycle (start/stop cycle) in which the idle operation mode and the standby operation mode make a circuit, and a second upper limit value as an upper limit value of a second power saving control cycle count showing an execution count of a second power saving control cycle (load/unload cycle) in which the idle operation mode and the power saving mode of either the low-speed rotation operation mode or the standby operation mode make a circuit;

wherein, during the course of controlling the operation mode of the storages belonging to each of the RAID groups (RG), the controller measures the first power saving control cycle count and the second power saving control cycle count; and

for storages of the RAID groups (RG) in the idle operation mode, selects the low-speed rotation operation mode or the standby operation mode in order to deconcentrate the increase of the first power saving control cycle count in relation to the first upper limit value and the increase of the second power saving control cycle count in relation to the second upper limit value in the RAID group.

8. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG), and the plurality of RAID groups (RG) have a plurality of logical devices (LDEV) as virtual logical areas configured therein;

wherein, during the course of controlling the operation mode of storages belonging to each of the RAID groups (RG), the controller measures the access interval to each of the logical devices (LDEV) belonging to each of the RAID groups (RG), and measures the frequency in which the measured access interval exceeds a prescribed setting time or the cumulative count in which the measured access interval exceeds the setting time, sets the logical device (LDEV) with the maximum frequency or cumulative count as a relocation target logical device (LDEV), selects a relocation destination RAID group among RAID groups (RG) other than the RAID group to which the relocation target logical device (LDEV) belongs, and moves the relocation target logical device (LDEV) to an unused area of the relocation destination RAID group.

9. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG), and the plurality of RAID groups (RG) have a plurality of logical devices (LDEV) as virtual logical areas configured therein;

wherein, during the course of controlling the operation mode of storages belonging to each of the RAID groups (RG), the controller measures the access interval to each of the logical devices (LDEV) belonging to each of the RAID groups (RG), and measures the frequency in which the measured access interval exceeds a prescribed setting time or the cumulative count in which the measured access interval exceeds the setting time, sets the logical device (LDEV) with the maximum frequency or cumulative count as a relocation target logical device (LDEV), selects a relocation destination RAID group among RAID groups (RG) other than the RAID group to which the relocation target logical device (LDEV) belongs, transfers information concerning the relocation target logical device (LDEV) and the relocation destination RAID group to a management terminal via a network, and, based on instructions from the management terminal, relocates the relocation target logical device (LDEV) to the relocation destination RAID group.

10. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG), and the plurality of RAID groups (RG) have a plurality of logical devices (LDEV) as virtual logical areas configured therein;

wherein, during the course of controlling the operation mode of storages belonging to each of the RAID groups (RG), the controller measures the access interval to each of the logical devices (LDEV) belonging to each of the RAID groups (RG), and measures the frequency in which the measured access interval exceeds a prescribed setting time or the cumulative count in which the measured access interval exceeds the setting time, additionally measures a wear-out rate showing the ratio of the measured value of the control cycle count and the upper limit value for each of the RAID groups (RG), sets the logical device (LDEV) with the maximum frequency or cumulative count as a relocation target logical device (LDEV), selects a RAID group (RG) in which the measured value of the wear-out rate was the smallest as a relocation destination RAID group among RAID groups (RG) other than the RAID group to which the relocation target logical device (LDEV) belongs, sets one of the logical devices (LDEV) belonging to the relocation destination RAID group as a replacement target logical device (LDEV), relocates the relocation target logical device (LDEV) to the relocation destination RAID group, and relocates the replacement target logical device (LDEV) to the RAID groups (RG) to which the relocation target logical device (LDEV) used to belong.

11. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG), and the plurality of RAID groups (RG) have a plurality of logical devices (LDEV) as virtual logical areas configured therein;

wherein, during the course of controlling the operation mode of storages belonging to each of the RAID groups (RG), the controller measures the access interval to each of the logical devices (LDEV) belonging to each of the RAID groups (RG), and measures the frequency in which the measured access interval exceeds a prescribed setting time or the cumulative count in which the measured access interval exceeds the setting time, additionally measures a wear-out rate showing the ratio of the measured value of the control cycle count and the upper limit value for each of the RAID groups (RG), sets the logical device (LDEV) with the maximum frequency or cumulative count as a relocation target logical device (LDEV), selects a RAID group (RG) in which the measured value of the wear-out rate was the smallest as a relocation destination RAID group among RAID groups (RG) other than the RAID group to which the relocation target logical device (LDEV) belongs, and moves the relocation target logical device (LDEV) to an unused area of the relocation destination. RAID group.

12. The storage system according to claim 1,

wherein the plurality of storage are divided into a plurality of RAID groups (RG),

wherein during the course of controlling the operation mode of storages belonging to each of the RAID groups (RG), the controller measures each of the control cycle counts for each of the RAID groups (RG);

and

for storages of the RAID groups (RG) in the idle operation mode, selects a power saving operation mode among the plurality of power saving operation modes based on the measured value for deconcentrating the increase in the measured value of each of the control cycle counts in relation to each of the upper limit values in the RAID group.

13. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG), and the plurality of RAID groups (RG) have a plurality of logical devices (LDEV) as virtual logical areas configured therein;

wherein, during the course of controlling the operation mode of storages belonging to each of the RAID groups (RG), the controller measures the access interval for accessing the logical devices (LDEV) belonging to each of the RAID groups (RG), and measures the frequency in which the measured access interval exceeds a prescribed setting time or the cumulative count in which the measured access interval exceeds the setting time, sets the logical device (LDEV) with the maximum frequency or cumulative count as a relocation target logical device (LDEV), and moves the relocation target logical device (LDEV) to semiconductor memory storages configuring a RAID group that is different from the plurality of RAID groups (RG).

14. The storage system according to claim 1,

wherein the plurality of storages are divided into a plurality of RAID groups (RG);

wherein, during the course of controlling the operation mode of storages belonging to each of the RAID groups (RG), the controller measures a control cycle count of each of the RAID groups (RG), calculates the incremental count of the measured value of the control cycle count as the frequency per day based on the measured value, estimates the life of each of the RAID groups (RG) from the time obtained by dividing the difference between the upper limit value set for the control cycle count and the measured value of the control cycle by the calculated frequency, and the current date, and displays the estimated result on a management screen of a management terminal via a network.