JP2000020447A - Storage subsystem - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide exclusive control by specifying a bus since exclusive control is disabled and data are destroyed by the disability of path specification concerning a peripheral device having plural paths (port, controller and LU) connected to the bus interface of an SCSI bus or the like in the system with no mount concept. SOLUTION: On the side of a disk array 100, controllers 140A-140B have the ownership corresponding to respective internal LU and the controllers 140A-140B successively number the LU having the ownership in the ascending order from '0' and respond to an SCSI command such as Inquiry issued from the host to the respective LU so that any bus can be specified. Therefore, in the case of connecting the peripheral device to the system with no mount concept, by specifying the path on the side of device, a mount can be provided.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer in which each device such as SCSI has a unique device ID and LUN, and is mutually connected by an interface for exchanging data, commands, messages, etc. with other devices. The present invention relates to a system and an external storage device having a plurality of ports, a plurality of controllers, and a plurality of LUs.

[0002]

2. Description of the Related Art Conventional computer system peripheral devices include, for example, ANSI (American Standards Association)
SI X3.131-1986 "Small Computer System
As described in “Interface (SCSI)”, each device is allowed to have a unique device ID and up to eight LUs.

FIG. 2 is a diagram showing a configuration of a storage subsystem having four LUs. Each drive is connected to two controllers 140A and 140B, and each controller is connected to the host PC / WS via four ports.
It is connected to the. In such a system, the host P
All the LUs can be seen from the C / WS 200 through the controller 140A, and all the LUs can be similarly seen from the host PC / WS 201 through the controller 140B. The plurality of drives 180 constitute a disk array, and the area of each drive is divided into LU0 to LU3 areas.

In a system as shown in FIG. 2, when two host PCs / WSs access the same LU, data destruction occurs in this LU.
Is performing exclusive control on the drive. Host P
When the C / WS operates on an OS (Operating System) that requires a necessary device (drive) to be incorporated in the system by specifying a drive by specifying a path by a mount command, as shown in FIG. PC
/ WS200 transfers LU0 and LU2 to host PC / W
Exclusive control can be performed by S201 mounting LU1 and LU3.

However, when the host PC / WS does not need to execute a mount command to incorporate a device into the system, that is, when the host PC / WS operates on an OS without a mount concept, the same host PC / WS receives the same command from a plurality of host PCs / WS. You can see the drive (LU). Therefore, exclusive control cannot be performed if there is a path that can access the same LU as shown in FIG. 4, so that writing to the same LU is possible from each path, leading to data destruction.

In the technology disclosed in Japanese Patent Application Laid-Open No. 5-265914, a path is specified by assigning a device ID to each LU (partition) on the device side. However, in a device having a plurality of controllers and a plurality of ports, a path cannot be specified by the above-described conventional technique in which a device ID is assigned to a partition.

[0007]

Conventionally, in a system having no mount concept, in a peripheral device having a plurality of ports, a plurality of controllers, and a plurality of LUs, a path (port, controller, LU) designated by the device side to a host is used. Could not be identified. For this reason, exclusive control and the like cannot be performed, and LUs being used by other hosts can be accessed, leading to data destruction. Therefore, it is necessary to realize exclusive control by specifying a path.

[0008]

In order to solve the above-mentioned problems, each controller or each port is given LU ownership, and the controller assigns LUNs from 0 to LUN in ascending order to the LU having ownership. By assigning and responding to a SCSI command such as Inquiry issued from the host side, the path (port, controller, L
U) is specified. By responding only to the LU having the owner right, it is possible to prevent the LU without the owner right from being shown to the specific host in a pseudo manner.

Further, according to the present invention, when the two controllers are operating normally, each controller responds only to the LU having the ownership, and when one of the controllers fails, an abnormal controller is returned. Has a configuration in which a normal controller can respond to an LU having ownership.

[0010]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a conceptual diagram showing one embodiment of a computer system according to the present invention. In the present embodiment,
As an example of the bus interface, a SCSI bus will be described as an example, but it is needless to say that the present invention can be applied to any bus interface such as a Fiber Channel.

In this embodiment, the storage subsystem 10
0 is a host PC with multiple OSs without a mount concept
/ WS200-201 and SCSI cable 120A-
B is connected. Each of the plurality of host PCs / WSs is, for example, an OS processing device without a mount concept, such as a personal computer or a workstation.

The storage subsystem 100 controls one or more storage devices 180 (which may be logically divided), such as a hard disk device, and controls transmission and reception of information between the storage devices and the outside. It is composed of a plurality of controllers 140A-B having a redundant configuration for the purpose of reliability and the like, a plurality of ports 130A-D serving as an information interface with the outside, and the like. An operation panel for inputting control information and a register for registering an LU accessible by each controller are connected to the controllers 140A and 140B. When the storage device 180 forms a disk array, the configuration is as shown in FIG. Even when a plurality of storage devices constitute a disk array, they appear as one storage device to the host PC / WS, and therefore are shown as one logically divided storage device in FIG.

The controller operates an operation panel 160.
The control of the device configuration and operation is performed by the control information set via the. By setting an accessible LU for each controller and each port using the operation panel 160, the LU is registered in the register 170. At this time, the following three combinations are conceivable as combinations of settings that can be specified. (1) Only the device ID can be specified. (2) Only LU can be specified. (3) A combination of ID and LU can be specified.

In FIG. 1, LU0 is provided to the controller 140A.
LU2 are assigned to the host PC / WS200 connected to the controller 140A, and LU0 and LU2 are set as LU0 and LU1, respectively.
LU1 and LU3 are assigned to the host PC / WS 201 connected to the controller 140B.
This is an example of a response when U3 is set as LU0 and LU1, respectively.

FIG. 5 shows another embodiment according to the present invention.
In this example, the controller 140A is assigned ownership of LU0 and LU2, and the controller 140B is assigned ownership of LU1 and LU3.
/ WS in response to a command.

FIG. 6 shows another embodiment according to the present invention.
In this example, a device ID is set for each port, and one LU is assigned to each port. For example, the controller 140A has a port 1 having ID1.
LU0 assigned to 30A, port 13 with ID2
LU2 is assigned to 0B. Controller 1
40B allocates LU2 to the port 131A having ID3, and allocates LU3 to the port 131B having ID4.

FIG. 7 shows another embodiment according to the present invention.
In this example, one device ID and multiple L
U is assigned.

FIG. 9 is a conceptual diagram showing an example of the configuration of a register table for storing LUNs assigned to each port of each controller set from the operation panel.
The vertical direction indicates each port of each controller, and the horizontal direction is the LUN for each port. The entry at the vertical / horizontal intersection is the actual LUN inside the disk array.

FIG. 8 shows another embodiment according to the present invention.
When a failure occurs, when a plurality of controllers are configured, when another controller failure is detected, the controller is closed, and the ID and LU assigned to the failed controller are automatically taken over by the normal controller. You can also. This enables automatic replacement in the event of a failure, and replaces (recovers) the failed controller.
In the meantime, it can be operated without stopping the system. For example, when a failure occurs in the controller 140A, the normal controller 140B automatically takes over the ID and LU assigned to the controller 140A and responds.

FIG. 10 shows another embodiment according to the present invention. An external input from the operation panel 160 allows the host P
It is also possible to set whether or not to respond to a command issued from the C / WS. As a result, access is possible at the time of installation or failure, but it is possible to disable access at the time of normal use.

[0022]

As described above, according to the present invention, by allocating a plurality of LUs to each controller and each port, it is possible to specify the LU corresponding to the allocated LUN. As a result, even in a system without a mount concept, the path (port, controller, LU)
Can be specified, the mount path can be specified on the device side, exclusive control is realized, and data destruction can be prevented.

Further, even when a controller fails, another controller takes over the LU on the failed controller side.
High reliability of the system can be realized.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram illustrating an example of a configuration of a storage subsystem according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram illustrating an example of a configuration of a storage subsystem according to an embodiment of the present invention.

FIG. 3 is a conceptual diagram showing an example of a configuration in a system having a mount concept.

FIG. 4 is a conceptual diagram showing an example of a configuration in a system without a mount concept.

FIG. 5 is a conceptual diagram illustrating an example of a configuration of a storage subsystem according to an embodiment of the present invention.

FIG. 6 is a conceptual diagram illustrating an example of a configuration of a storage subsystem according to an embodiment of the present invention.

FIG. 7 is a conceptual diagram illustrating an example of a configuration of a storage subsystem according to an embodiment of the present invention.

FIG. 8 is a conceptual diagram illustrating an example of a configuration of a storage subsystem according to an embodiment of the present invention.

FIG. 9 is a conceptual diagram showing a setting example of a disk array side configuration of a storage subsystem according to an embodiment of the present invention.

FIG. 10 is a conceptual diagram illustrating an example of a configuration of a storage subsystem according to an embodiment of the present invention.

[Explanation of symbols]

200-201: Host PC / WS 200A, 201A: Host-side interface port 100: Storage subsystem 160: Operation panel 140A, 140B: Controller 130A-130D: Port A-Port D 131A-131D: Port A-Port D 120A- 120B: SCSI cable A to SCSI cable B 170: Register 180: Storage device 190: LU

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Tetsuzo Kobashi 2880 Kozu, Kozuhara-shi, Kanagawa Prefecture Storage Systems Division, Hitachi, Ltd. F-term in the system division (reference) 5B014 HA02

Claims (4)

[Claims] A storage device having a first storage area for storing data for a first host device and a second storage region for storing data for a second host device; A storage subsystem comprising: a first higher-level device and a first controller connected to the storage device; and a second controller connected to the second higher-level device and the storage device; A storage subsystem, wherein a controller accesses only the first storage area, and the second controller controls so as to access only the second storage area. 2. The storage device according to claim 1, wherein said storage device has a plurality of said first storage areas, and said plurality of storage areas are given consecutive logical unit numbers. sub-system. 3. A plurality of first storage areas for storing data for a first host device and a plurality of second storage areas for storing data for a second host device, A storage device connected to the first controller and the second controller; a first controller connected to the first host device via a plurality of first ports; and a plurality of storage devices connected to the second host device And a second controller connected through a second port of the first storage area, wherein the first controller is connected to each of the plurality of first ports in the first storage area. The storage subsystem, wherein one storage area is allocated, and the second controller allocates one storage area of the second storage area to each of the plurality of second ports. 4. A storage device having first and second storage areas for storing data for a host device, and a first controller and a second controller connected to the host device and the storage device. Wherein the first controller accesses only the first storage area, the second controller accesses only the second storage area, and a failure occurs in the first controller. When an error occurs, the second controller accesses the first storage area, and when a failure occurs in the second controller, the first controller accesses the second storage area. Storage subsystem.