Distributed storage system, and data-access method therefor

• US 9,940,042 B2
• Filed: 09/06/2013
• Issued: 04/10/2018
• Est. Priority Date: 09/06/2013
• Status: Active Grant

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Agetsuma, Masakuni
• Primary Examiner(s): Thai, Tuan
• Assistant Examiner(s): Waddy, Jr., Edward
• Application Number:
  US14/900,791
• Publication Number:
  US 20160139841A1
• Time in Patent Office: 1,677 Days
• Field of Search: 711162, 711E12066, 707E17032, 707611, 707624, 707827, 707999201, 1 1
• CPC Class Codes: G06F 11/14 Error detection or correction of the data by redundancy in operation G06F11/16 takes precedence
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 16/178 Techniques for file synchronisation in file systems
• CPC Class Codes: G06F 3/06 Digital input from, or digital output to, record carriers , e.g. RAID, emulated record carriers or networked record carriers
• CPC Class Codes: G06F 3/0611 in relation to response time
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]

Abstract

Each of a plurality of storage systems in a distributed storage system is provided with: a strong-consistency-control processor which controls data synchronization that ensures consistency of duplicated data; a weak-consistency-control processor which controls data synchronization that does not necessarily ensure consistency of the duplicated data; an access-switching processor which determines a redundancy number corresponding to the number of storage systems having duplicated data stored thereon, selects in accordance with the characteristics of the duplicated data and the position of a terminal issuing the I/O request, storage systems of a number corresponding to the redundancy number, and selectively determines, as the control for the data synchronization, one from among the strong-consistency-control processor, the weak-consistency-control processor, and consistency control unnecessary; and an I/O processor which executes, on the basis of the determined control for the data synchronization, the I/O request for the duplicated data in the selected storage systems.

Claims

1. A distributed storage system including a plurality of storage systems coupled with one another via a network, a first storage system among the plurality of storage systems comprising:

a strong-consistency-control processing unit which, during access to duplicated data stored in the distributed storage system, controls data synchronization that ensures the consistency of the duplicated data that the accessed duplicated data is the latest;

a weak-consistency-control processing unit which, during access to the duplicated data, controls data synchronization that does not necessarily ensure the consistency of the duplicated data;

an access-switching processing unit which;

in response to an I/O request for the duplicated data, determines, in accordance with policies predetermined in accordance with characteristics of the duplicated data, a redundancy number corresponding to the number of second storage systems to have the duplicated data stored thereon;

selects, from among the plurality of storage systems, in accordance with the characteristics of the duplicated data and a position of a terminal issuing the I/O request, second storage systems of a number corresponding to the redundancy number; and

selectively determines via an access-switching policy and a policy template that correspond to the I/O request, as the control for the data synchronization, one from among a strong-consistency-control scheme, a weak-consistency-control scheme, and an unnecessary consistency control scheme, wherein the policy templates include information regarding the consistency used for determining the control for the data synchronization, the access response, and the redundancy number;

wherein the unnecessary consistency control scheme, consistency and redundancy of the I/O target data is unnecessary, because the I/O processing can be performed on the respective storage systems without executing consistency processing; and

an I/O processing unit which executes, on the basis of the determined control for the data synchronization, the I/O request for the duplicated data in the selected second storage systems.

2. The distributed storage system according to claim 1,

wherein the second storage systems, which are selected in accordance with the redundancy number, include storage systems in the vicinity of the position of the terminal.

3. The distributed storage system according to claim 2,

wherein each of the second storage systems is selected on the basis of at least one of its geographic distance or its distance on the network from the first storage system while its access response is taken into reflection.

4. The distributed storage system according to claim 3,

wherein the strong-consistency-control processing unit uses quorum for checking whether a majority of the duplicated data, which are to be stored in the first storage system and the second storage systems, has already been written or not in order to ensure the consistency of the duplicated data.

5. The distributed storage system according to claim 4,

wherein the first storage system is one of the second storage systems.

6. The distributed storage system according to claim 5,

wherein the characteristics of the duplicated data are judged on the basis of the path name of a directory which stores the duplicated data included in the I/O request.

7. The distributed storage system according to claim 6,

wherein the policies include conditions regarding the positions and credibilities of the second storage systems which store the duplicated data, and policy templates to be used, and

the policy templates include information regarding the consistency used for determining the control for the data synchronization, the access response, and the redundancy number.

8. A data access method in a distributed storage system including a plurality of storage systems coupled with one another via a network,

wherein a first storage system among the plurality of storage systems performs the steps of:

determining, in response to an I/O request for the duplicated data and in accordance with policies predetermined in accordance with characteristics of the duplicated data, a redundancy number corresponding to a number of second storage systems to have the duplicated data stored thereon;

selecting, from among the plurality of storage systems, in accordance with the characteristics of the duplicated data and a position of a terminal issuing the I/O request, second storage systems of a number corresponding to the redundancy number;

selectively determining via an access-switching policy and a policy template that correspond to the I/O request, as the control for the data synchronization, one from among (1) strong consistency control scheme which, during access to the duplicated data stored in the distributed storage system, controls data synchronization that ensures the consistency of the duplicated data that the accessed duplicated data is the latest, (2) weak consistency control scheme which, during access to the duplicated data, controls data synchronization that does not necessarily ensure the consistency of the duplicated data, and (3) an unnecessary consistency control scheme, wherein the policy templates include information regarding the consistency used for determining the control for the data synchronization, the access response, and the redundancy number;

wherein the unnecessary consistency control scheme, consistency and redundancy of the I/O target data is unnecessary, because the I/O processing can be performed on the respective storage systems without executing consistency processing; and

executing, on the basis of the determined control for the data synchronization, the I/O request for the duplicated data in the selected second storage systems.

9. The data access method according to claim 8,

wherein the second storage systems, which are selected in accordance with the redundancy number, include storage systems in the vicinity of the position of the terminal.

10. The data access method according to claim 9,

wherein each of the second storage systems is selected on the basis of at least one of its geographic distance or its distance on the network from the first storage system while its access response is taken into reflection.

11. The data access method according to claim 10,

wherein the strong consistency control uses quorum for checking whether a majority of the duplicated data, which is to be stored in the first storage system and the second storage systems, has already been written or not in order to ensure the consistency of the duplicated data.

12. The data access method according to claim 11,

wherein the first storage system is one of the second storage systems.

13. The data access method according to claim 12,

wherein the characteristics of the duplicated data are judged on the basis of the path name of a directory which stores the duplicated data included in the I/O request.

14. The data access method according to claim 13,

wherein the policies include conditions regarding the positions and credibilities of the second storage systems which store the duplicated data, and policy templates to be used, and the policy templates include information regarding the consistency used for determining the control for the data synchronization, the access response, and the redundancy number.