STORAGE DEVICE

• US 20170083417A1
• Filed: 07/25/2014
• Published: 03/23/2017
• Est. Priority Date: 07/25/2014
• Status: Active Grant

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Kawaguchi, Tomohiro, SIMOZONO, Norio, Saito, Hideo, Mizuno, Makio
• Granted Patent: US 10,296,429 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 11/0727 in a storage system, e.g. in a DASD or network based storage system drivers for digital recording or reproducing units G06F3/06; circuits for error detection or correction within digital recording or reproducing units G11B20/18; for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS], H04L67/1097
• CPC Class Codes: G06F 11/0757 by exceeding a time limit, i.e. time-out, e.g. watchdogs
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06F 11/1658 Data re-synchronization of a redundant component, or initial sync of replacement, additional or spare unit
• CPC Class Codes: G06F 11/201 between storage system components
• CPC Class Codes: G06F 11/2092 Techniques of failing over between control units
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 13/00 Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units interface circuits for specific input/output devices G06F3/00 ; multiprogram control therefor G06F9/46; multiprocessor systems G06F15/16
• CPC Class Codes: G06F 13/10 Program control for peripheral devices G06F13/14 - G06F13/42 take precedence
• CPC Class Codes: G06F 2201/805 Real-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 3/0683 Plurality of storage devices

Abstract

A storage subsystem comprises one or more volumes, and multiple nodes having multiple control packages interconnected via an intra-node communication path, wherein the control packages of different nodes are interconnected via an inter-node communication path having a lower transmission path capacity than the intra-node communication path. When the host computer accesses a volume, access is enabled via any of at least two or more control packages out of the multiple control packages, and the priority for issuing access requests to the relevant volume is determined in each of the control packages. When the storage subsystem detects failure, it changes the priorities determined for the control packages according to the failure occurrence portion, and notifies the same to the host computer. The host computer determines the control package being the issue destination of the access request based on the notified priority.

Claims

1. A storage subsystem comprising multiple control nodes for processing requests from a host computer, and one or more volumes;

wherein

each of the multiple control nodes has multiple control packages each having a processor, a memory, and a port for receiving accesses from the host computer;

the respective control packages within the node are mutually interconnected via a first communication path;

the control package is interconnected with a control package belonging to a node that differs from the node to which said control package belongs via a second communication path having a lower performance than the first communication path;

the host computer can access the volume via any one of at least two or more control packages out of the multiple control packages, and each of the at least two or more control packages has a priority determined for issuing an access request from the host computer to the volume; and

when the storage subsystem detects failure, it changes the priority determined for the control package according to a failure occurrence section, and notifies the priority having been changed to the host computer.

2. The storage subsystem according to claim 1, wherein

the control package being the owner is determined in each of said one or more volumes, which is configured to process an I/O request arriving from the host computer to the volume;

and

when the storage subsystem detects failure, it changes the control package becoming the owner of the volume according to a failure occurrence section, changes the priority of the control package becoming the owner of the volume to a highest priority, and notifies the changed priority to the host computer.

7. The storage subsystem according to claim 2, wherein

the storage subsystem determines the priority of the control package being the owner of the volume to be the highest, the priority of the control package belonging to a same node as the control package being the owner of the volume to be the second highest, and the priority of a control package belonging to a different node as the control package being the owner of the volume to be the third highest;

when not all the control packages of the node to which the control package being the owner of the volume belongs can receive an access request, the storage subsystem sets the control package having the third highest priority to receive an access request to said volume from the host computer; and

the control package having the third highest priority transfers the access request via the second communication path to the control package being the owner of the volume.

8. The storage subsystem according to claim 2, wherein

a part of an area of the memory in the control package is configured as a cache area for caching the data of the volume;

and

when a write request to the volume is received from the host computer, a first control package being the owner of the volume stores a write data in the cache area within the first control package, and causes a second control package determined as a sub-owner having a role to store a replica of the write data to the volume to store the write data in the cache area within the second control package.

9. The storage subsystem according to claim 8, wherein

when the storage subsystem detects that failure has occurred to the control package being the owner of the volume;

if multiple normal control packages do not exist in the node to which the control package where failure has occurred belongs, the control package belonging to a node that differs from the node to which the control package where failure has occurred belongs is set as the owner of the volume; and

dirty data stored in the cache area within the second control package determined as a sub-owner of the volume is destaged to a storage device constituting the volume.

10. The storage subsystem according to claim 8, wherein

when the control package not being the owner of the volume receives a given threshold or more of access requests to the volume from the host computer;

the storage subsystem changes the control package not being the owner of the volume to be the owner of the volume;

changes the control package belonging to a same node as the node to which the control package not being the owner of the volume belongs as a sub-owner of the volume; and

as a result of the change, when the control package having been the owner of the volume has become neither the owner nor the sub-owner of the volume, destages a dirty data stored in the cache area within the control package having been the owner of the volume to the storage device constituting the volume.

11. The storage subsystem according to claim 8, wherein

when the storage subsystem detects that failure has occurred to the control package being the owner of the volume;

if multiple normal control packages do not exist in the node to which the control package where failure has occurred belongs, the storage subsystem sets the control package belonging to a node that differs from the node to which the control package where failure has occurred belongs as the owner of the volume; and

copies a dirty data stored in the cache area within the control package having been the sub-owner of the volume to the cache area within the control package set to become the owner of the volume.

12. The storage subsystem according to claim 8, wherein

when a given threshold value or more access requests from the host computer to the volume is received by the control package not being the owner of the volume, the storage subsystem

changes the control package not being the owner of the volume to become the owner of the volume;

changes the control package belonging to a same node as the node to which the control package changed as the owner of the volume belongs to become a sub-owner of the volume;

as a result of the change, if the control package having been the owner of the volume is changed to become the sub-owner of the volume, the subsystem copies a clean data out of the data corresponding to the volume within the control package having been the owner of the volume to the cache area within the control package being changed to be the owner of the volume; and

as a result of the change, if the control package having been the owner of the volume has become neither the owner nor the sub-owner of the volume, it copies all the data corresponding to the volume within the control package having been the owner of the volume to the cache area within the control package being changed to be the owner of the volume.

3. The Storage subsystem according to claim 3, wherein

a different control package is changed to become the owner of the volume for each case, respectively, when failure of the control package being the owner of the volume is detected, and when a failure of access path between the control package being the owner of the volume and the host computer is detected.

4. The storage subsystem according to claim 3, wherein

when it is detected that failure has occurred to the control package being the owner of the volume;

if multiple normal control packages do not exist in the node to which said control package where failure has occurred belongs, the control package belonging to a node that differs from said node to which the control package where failure has occurred belongs is set as the owner of the volume.

5. The storage subsystem according to claim 3, wherein

when it is detected that failure has occurred to the access path between the control package being the owner of the volume and the host computer;

the control package belonging to the same node as said node to which the control package where failure has occurred belongs is set as the owner of the volume.

6. The storage subsystem according to claim 5, wherein

when the control package not being the owner of the volume receives access requests from the host computer to said volume for a given threshold value or greater;

it is determined that failure has occurred to the access path between the control package being the owner of the volume and the host computer.

13. A computer system comprising a host computer, and a storage subsystem having multiple control nodes for processing requests from a host computer and one or more volumes;

wherein

each of the multiple control nodes has multiple control packages each having a processor, a memory, and a port for receiving accesses from the host computer;

the respective control packages within the node are mutually interconnected via a first communication path;

the control package is interconnected with respective said control package belongs via a second communication path having a lower transmission path capacity than the first communication path;

the host computer can access the volume via any one of at least two or more control packages out of the multiple control packages, and each of the at least two or more control packages has a priority determined for issuing an access request from the host computer to the volume;

when the storage subsystem detects failure, it changes the priority determined for the control package according to a failure occurrence section, and notifies the priority having been changed to the host computer; and

the host computer selects the control package to be an issue destination of the access request to the volume based on the priority notified from the storage subsystem.

14. The computer system according to claim 13, wherein

the host computer issues an access request with respect to the volume to the control package having the highest priority.

15. The computer system according to claim 14, wherein

the control package being the owner is determined in each of said one or more volumes in the storage subsystem, which is configured to process an I/O request arriving from the host computer to the volume;

the storage subsystem determines the priority of the control package being the owner of the volume to be the highest, and the priority of at least one or more control packages belonging to a same node as the control package being the owner of the volume to be the second highest; and

when the host computer cannot issue an access request regarding the volume to the control package having the highest priority, it issues an access request regarding the volume to the control package having the second highest priority.