Distributed storage system and method

• US 9,690,658 B2
• Filed: 03/13/2015
• Issued: 06/27/2017
• Est. Priority Date: 03/29/2014
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Kato, Jun
• Primary Examiner(s): Chaudry, M. Mujtaba
• Application Number:
  US14/656,756
• Publication Number:
  US 20150278018A1
• Time in Patent Office: 837 Days
• Field of Search: None
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 2211/1028 Distributed, i.e. distributed RAID systems with parity

Abstract

A distributed storage system includes a plurality of storage devices and an information processing devices including a storage unit that is accessed more preferentially than the plurality of storage devices, and a processor. The processor deletes the data that can be restored from the storage unit when data that belongs to data stored in a first storage device and which can be restored is stored in the storage unit. The processor writes data that belongs to the data stored in the first storage device and which will not be restored to the storage unit. The processor switches the first storage device to the power-off state after writing the data. The processor reads the data that has been stored in the storage unit and which will not be restored when a request to read the data that will not be restored is given.

Claims

1. A distributed storage system, comprising:

a plurality of storage devices; and

an information processing device;

wherein

the information processing device includes;

a cache; and

a processor that executes a process including

deleting first data from the cache, when the first data is stored in the cache and a first storage device of the plurality of storage devices, and the first data being capable of being restored by using second data and a parity, the second data and the parity being distributed and stored in a second storage device of the plurality of storage devices that is in the power-on state,

writing third data to the cache, the third data being stored in the first storage device and being not restored even by using the second data and the parity that are distributed and stored in the second storage device that is in the power-on state,

switching the first storage device to the power-off state after writing the third data to the cache, and

reading the third data that has been stored in the cache when a request to read the third data is given when the first storage device is in the power-off state.

2. The distributed storage system according to claim 1, wherein the process executed by the processor further including

acquiring the first data that can be restored by using the second data and the parity that are distributed and stored in the second storage device that is in the power-on state when a request to read the first data is given when the first storage device is in the power-off state, and

the reading processing that is performed by the processor further includes processing to read the first data acquired by the acquiring processing.

3. The distributed storage system according to claim 1, wherein the writing processing that is performed by the processor further includes processing to write data having a high possibility of being used to the cache when writing the third data to the cache.

4. The distributed storage system according to claim 1, wherein

the writing processing that is performed by the processor further includes processing to write the third data to the second storage device that is in the power-on state when a request to update the third data that has been stored in the first storage device is given when the first storage device is in the power-off state, and further includes processing to write data having a high possibility of being used from among fourth data stored in a third storage device of the plurality of storage devices that is in the power-on state to the cache after the third data that has been stored in the cache is deleted,

the deletion processing that is performed by the processor further includes processing to delete the third data before updating that has been stored in the cache after the third data is written to the second storage device in the power-on state, and

the reading processing that is performed by the processor further includes processing to read the third data that has been stored in the second storage device when a request to read the third data is given when the first storage device is in the power-off state and the third data has been written to the second storage device.

5. The distributed storage system according to claim 1, wherein the process executed by the processor further including

acquiring the first data that is identified by a data identifier stored in the cache when a third storage device of the plurality of storage devices enters the power-on state when the first storage device is in the power-off state,

the writing processing that is performed by the processor further includes processing to write the data identifier of the first data to the cache after the first data is deleted from the cache by the deleting processing, to write the acquired first data to the cache when the first data is acquired by the acquiring processing, and to write the parity stored in second storage device in the power-on state to the third storage device,

the switching processing that is performed by processor further includes processing to switch the third storage device to the power-off state after the parity stored in second storage device that is in the power-on state is written to the third storage device,

the deleting processing that is performed by the processor further includes processing to delete the parity stored in second storage device that is in the power-on state from the second storage device after the third storage device is switched to the power-off state, and

the reading processing that is performed by the processor further includes processing to read the first data that has been stored in the cache when a request to read the first data that has been acquired is given.

6. The distributed storage system according to claim 1, wherein

the deleting processing that is performed by the processor further includes processing to delete the third data from the cache after the first storage device enters the power-on state, and

the writing processing that is performed by the processor further includes processing to write data having a high possibility of being used from among fourth data stored in a third storage device of the plurality of storage devices that is in the power-on state to the cache.

7. The distributed storage system according to claim 5, wherein

the writing processing that is performed by the processor includes processing to write the parity and the third data to an empty area of the second storage device that is in the power-on state, and

the switching processing that is performed by the processor includes processing to switch the third storage device to the power-on state when the empty area of the second storage device that is in the power-on state becomes lower than a predetermined amount.

8. The distributed storage system according to claim 1, wherein

the writing processing that is performed by the processor includes processing to write the parity and the third data to an empty area of the second storage device that is in the power-on state, and to write the third data that has been stored in the second storage device that is in the power-on state to the first storage device after the first storage device enters the power-on state,

the switching processing that is performed by the processor includes processing to switch the first storage device to the power-on state when the empty area of the second storage device that is in the power-on state becomes lower than a predetermined amount, and

the deleting processing that is performed by the processor includes processing to delete the third data that has been stored in the second storage device that is in the power-on state when the first storage device has entered the power-on state and when the third data has been written to the first storage device.

9. The distributed storage system according to claim 1, wherein the information processing device or the other information processing devices includes the cache or the processor or any combination thereof.

10. A storage device control method that is performed by a processor of an information processing device including a cache and connected to a plurality of storage devices so as to be capable of communication, the storage device control method comprising:

a process that is performed by the processor including

deleting first data from the cache, when the first data is stored in the cache and a first storage device of the plurality of storage devices, and the first data being capable of being restored by using second data and a parity, the second data and the parity being distributed and stored in a second storage device of the plurality of storage devices that is in the power-on state,

writing third data to the cache, the third data being stored in the first storage device and being not restored even by using the second data and the parity that are distributed and stored in the second storage device that is in the power-on state;

switching the first storage device into the power-off state; and

reading the third data that has been stored in the cache when a request to read the third data is given when the first storage device is in the power-off state.

11. The storage device control method according to claim 10, wherein the process that is performed by the processor further including

acquiring the first data that can be restored by using the second data and the parity that are distributed and stored in the second storage device that is in the power-on state when a request to read the first data is given when the first storage device is in the power-off state;

and

reading the acquired first data.

12. The storage device control method according to claim 10, wherein the process that is performed by the processor further including writing data having a high possibility of being used to the cache in the processing to write the third data to the cache.

13. A non-transitory computer-readable recording medium storing programs for causing a processor of an information processing device including a cache and connected to a plurality of storage devices so as to be capable of communication, to perform storage device control process, the storage device control process comprising:

a process that is performed by the processor including

deleting first data from the cache, when the first data is stored in the cache and a first storage device of the plurality of storage devices, and the first data being capable of being restored by using second data and a parity, the second data and the parity being distributed and stored in a second storage device of the plurality of storage devices that is in the power-on state,

writing third data to the cache, the third data being stored in the first storage device and being not restored by using the second data and the parity that are distributed and stored in the second storage device that is in the power-on state;

switching the first storage device into the power-off state; and

reading the third data that has been stored in the cache when a request to read the third data is given when the first storage device is in the power-off state.

14. The non-transitory computer-readable recording medium according to claim 13, wherein the process that is performed by the processor further including

acquiring the first data that can be restored by using the second data and the parity that are distributed and stored in the second storage device that is in the power-on state when a request to read the first data is given in the state when the first storage device is in the power-off state;

and

reading the acquired first data.

15. The non-transitory computer-readable recording medium according to claim 13, wherein the process that is performed by the processor further including writing data having a high possibility of being used to the cache in the processing to write the third data to the cache.