JPH10187358A - Disk array device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To inexpensively perform the lighting control of an LED that shows a failed disk drive by providing a failure detection LED controlling means which generates and refers to an ID management table that associates a certain disk drive with a disk drive which performs an LED light emitting control that corresponds to it. SOLUTION: When a certain disk drive 300 fails, a notification is sent to a disk interface controlling means controlling part through a disk interface controlling means 108. A failure detection LED controlling means notifies a RAID controlling part of a management ID that is related to the failed disk drive 300 based on an ID management table and issues an LED light emitting command through the disk interface controlling means controlling part. An LED controlling means 301 makes an LED 203 that corresponds to a managed disk that is connected via an LED wiring part 202 and a switch a 204 emit light. Thereby, the lighting control of an LED that shows a failed disk drive can inexpensively be performed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control method for a disk array system used as a storage system of a computer system.
The present invention relates to a technique for managing a disk device housing.

[0002]

2. Description of the Related Art Disk array systems use RAID (R)
It is called edundant Arrays of Inexpensive Disks, and has a configuration in which a plurality of disk devices are arranged in an array. This is a storage device in which reliability is improved by being added.

A host computer using a RAID as a storage device is often used for applications requiring continuous operation. Therefore, when a disk device in the RAID fails, the host computer or the RAID is turned off. There is a high need to replace disk devices without having to replace them. The replacement of the disk device while the power is turned on in this way is called hot swapping, and is realized by a disk case in which the disk device is stored. This disk case may be common to the host computer or may be different from the host computer.

Display means such as an LED for visually notifying a user of a failed disk device to be replaced is provided on the disk case, and means for detecting that a hot-line has been inserted or removed by the user. Is provided.

[0005] As a conventional technology of disk case management,
"SCSI Access" issued by CONNER, USA
ed Fault-Tolerant (SAF-TE)
Enclosures White Paper ". In the SAF-TE, the state management of the disk is realized by one SCSI device dedicated for management in the disk case. This SAF-TE uses an LED to indicate to a user a failed disk device,
The management-dedicated SCSI device controls the turning on / off of the LED in response to a request by the SCSI command from the host.

[0006]

According to the prior art,
Since the management of the disk status is realized by one management-dedicated SCSI device connected to the SCSI bus, when a RAID is constructed using the disk device, the means for controlling the RAID and the disk case must be connected. There is an effect that no special dedicated line is required.

[0007] However, the SCSI device requires a protocol analysis processor for control and a dedicated interface controller, and thus has a problem of high price.

Further, since the program for controlling the protocol analysis processor requires a large number of development steps, there is a problem that the system becomes expensive.

Further, a plurality of SCSs are provided on the disk case.
In the case where I is built in, there is a problem that a dedicated SCSI device for managing the number of SCSI is required to manage one disk case, resulting in a high-priced system.

A first object of the present invention is to connect a means for performing RAID control to a disk case without using a special dedicated line without using a dedicated device for disk management, and to provide an LED indicating a failed disk device. Lighting control at a low price.

A second object of the present invention is to connect a means for performing RAID control to a disk case without using a special dedicated line without using a dedicated device for disk management, and to provide an activity when replacing a disk device. It is to provide a line insertion / removal control at a low price.

[0012]

In order to achieve the first object, according to the present invention, a host computer is provided with a RAID control means and a RAID control program, a plurality of disk devices are stored in a disk case, and a host computer is provided. A computer system in which the computer and the disk case are connected by a disk interface bus that transfers data to and from the disk.
The disk device is provided with LED control means for controlling the LED emission of another disk device, and the disk case has
An LED corresponding to each of the plurality of disk devices and LED management disk determining means for uniquely determining a certain disk device and a disk device that performs control to emit the corresponding LED are provided. The RAID control program of the host computer includes: A failure detection LED control means for creating and referencing an ID management table for associating a disk device with a disk device that controls the corresponding LED to emit light;
The host computer is provided with management information holding means for storing the ID management table.

In order to achieve the second object, according to the present invention, a host computer is provided with a RAID control means and a RAID control program, a plurality of disk devices are stored in a disk case, and the host computer and the disk case are provided. A computer system in which the bodies are connected by a disk interface bus for transferring data to and from a disk. The disk device is provided with connection status recording means for monitoring and recording the connection status of the interface bus with other disk devices. The body includes a connection investigating means for indicating whether each of the plurality of disk devices is connected to the interface bus, and a disk device for monitoring whether or not a certain disk device is connected to the interface bus. A hot-swap management disk determining means for determining The RAID control program of the host computer is provided with a hot-swap controller for creating and referring to an ID management table for associating a disk device and a disk device for monitoring whether or not the disk device is connected to the interface bus. Is provided with management information holding means for storing the ID management table.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the present invention will be described with reference to the drawings. The first embodiment is for notifying a user of a failed disk device by emitting an LED from a nearby disk device.

FIG. 1 is a system configuration diagram of a first embodiment of the present invention. In FIG. 1, 100 is a host computer, 2
00 is a disk case for storing a disk device, 30
0-0 to 300-n are disk devices.

The host computer 100 includes a host CPU (central processing unit) 101, a cache 102 as a high-speed readable / writable memory, and a chipset 103 for controlling data transfer.
4 is connected. The chipset 103 is connected to a main memory 109 for storing data and programs, and a system bus 105. The main memory 109 has a RAID control program 110 to be described later, a management information holding unit 111 having an ID management table 130 as shown in FIG. 3, and a user area 112 for storing user data.

In the first embodiment, the RAID control means 106 is an expansion board type built in the host computer 100. The RAID control unit 106 includes the bus interface 107 and the disk interface control unit 10.
8 and is connected to the system bus 105 via the bus interface 107. The disk interface control means 108 and the bus interface 107 are connected, and the disk interface control means 108
It is connected to the disk case 200 via the disk interface bus 201. RA of the first embodiment
The ID control means 106 does not have a processor for RAID management and control.
The RAID control program in the control program 110 is executed by the host CPU 101 to manage and control the RAID.

FIG. 2 is a block diagram of the RAID control program 110 according to the first embodiment. The RAID control program 110 includes a request receiving unit 120 for receiving a request from an application, a request ending unit 121 for notifying the application of a request end, a RAID control unit 122 for managing and controlling RAID, and a disk device 300-0. ~
A disk interface control means control unit 123 for receiving a command issuance to 300-n and an end notification, and a failure detection for managing which LED should be lit to notify a user of a failed disk device inside the disk case 200 It comprises LED control means 124.

Returning to FIG. 1, the disk case 200 is
Disk interface bus 201 and bus connection unit 2
05-0 to 205n and switches a204-0 to 204
The LED wiring section 202 is connected between the switches a204-0 to 204-n, and the LED wiring section 202 is connected to one of the switches a204-0 to 204-n. The switches a 204-0 to 204-n and the LED wiring unit 202 are collectively referred to as an LED management disk determination unit 20.
6, the LED management disk determination means 206
, As described later, uniquely determines the relationship between the failed disk device 301 and the management disk that emits the LED 203 corresponding to the disk device 301. Switch a2
The other connection destination of 04-0 to 204-n is a switch a next to the LED wiring unit 202. In addition,
n is a natural number equal to or less than the number obtained by subtracting 1 from the maximum number of disk devices to which the disk interface bus 201 can be connected. Bus connection units 205-0 to 205-n
Are connected to the disk interface bus 201.

The disk devices 300-0 to 300-n are:
It is stored in the disk case 200, and includes LED control means 301-0 to 301-n for receiving an LED light emission command and controlling the light emission of the LED.

In a state where each of the disk devices 300-0 to 300-n is not stored in the disk case 200, the corresponding switches a204-0 to 204-n are operated by LEDs.
The wiring unit 202 is connected, and is connected to the next switch a. In FIG. 1, the switch a204-2
Is in this state.

When each of the disk devices 300-0 to 300-n is stored in the disk case 200, it is connected to the corresponding bus connection unit 205-0 to 205n so that data and control information can be exchanged with the host computer 100. And each of the disk devices 300-0 to 300-n
Sets the corresponding switches a204-0 to 204-n to L
Switching is performed so as to connect to EDs 203-0 to 203-n. In FIG. 1, the switch a204-0 and the like are in this state. In this state, the LED control means 301- in each of the disk devices 300-0 to 300-n.
0-301-n correspond to the corresponding switches a204-0-0
By being connected to 204-n as shown in FIG. 1, it is connected to the next switch a.

Therefore, the disk device 300- of FIG.
In the case where two disk devices 300 are stored consecutively, such as the disk device 300-1 and the disk device 300-1, the LED control means 301-0 inside the disk device 300-0 operates via the switch a204-1. , Next Disk Unit 3
It is in a state of being connected to the LED 203-1 corresponding to 00-1.

In the case where two disk devices 300 are stored in a state where no disk device is stored, as in the disk device 300-1 and the disk device 300-3 in FIG. LE inside 300-1
The D control means 301-1 is connected to the LEDs 203-3 corresponding to the two disk devices 300-3 via the switch a204-2 and the switch a204-3.

Similarly, even when two or more devices are spaced apart from each other, the switch a 204 in a place where no disk device is stored is connected to connect the LED wiring unit 202, so that a certain disk device 300 The LED control means 301 is connected to the switch a204 corresponding to the disk device which is stored closest to the wiring on the side where the LED control means 301 is connected to the LED wiring section 202.

The LED 203 connected to the LED control means 301 as described above can emit light by the LED control means 301.

The LED wiring section 202 includes a switch a 204
-0 to 204-n, respectively.
Since the switch a204-n at the end is connected to the switch a204-0, no matter which of the disk devices 300-0 to 300-n is stored in the disk case 200, the switch a204-n is connected to the next switch a204. The conditions will be the same.

The LEDs 203-0 to 203-n are:
It is sufficient if the arrangement is such that the end numbers correspond to the same disk devices 300-0 to 300-n. Further, in the first embodiment, for convenience of explanation, the disk interface bus 2 in the disk case 200 will be described.
It is assumed that the ID (identification number) of each disk device 300 on 01 is the same as the end number of each disk device 300-0 to 300-n.

Next, an operation for notifying the user which disk device 300 has failed will be described. In the first embodiment, another disk device 300 emits an LED 203 corresponding to the failed disk device 300 to notify the user of the failed disk device 300 to be replaced. In the example of FIG. 1, the LED 203-1 having the same end number corresponds to the disk device 300-1. In the first embodiment, the disk device 300 that causes the LED 203 to emit light is called a management disk, and its ID is
Is referred to as a management ID. Also, the disk device 300 that notifies the user of the failure by the management disk is called a managed disk, and its ID is assigned to the managed I / O.
D. Each disk device 300 stored in the disk case 200 becomes a management disk or a managed disk.

In the first embodiment, as described in the above system configuration, the LED control means 301 connected to the LED wiring section 202 via the switch a204.
The management disk and the managed disk are determined based on the relationship between the management disk and the LED 203. The disk device 300 having the connected LED control means 301 becomes a management disk,
The disk device 300 corresponding to D203 is a managed disk. In the example of FIG. 1, the disk device 300-0
Is a management disk, the disk device 300-1 corresponding to the LED 203-1 is a managed disk. By determining the management disk and the managed disk in this manner, the management disk can emit the LED 203 corresponding to the managed disk.

First, a method of creating an ID management table 130 for associating a managed disk with a managed disk will be described. At the time of system initialization, the RAID control unit 122 sends a command for confirming whether or not a disk device is actually stored to all IDs on the disk interface bus 201 via the disk interface control unit control unit 123. Issue. Based on the result, the failure detection LED control unit 124 creates the ID management table 130. Disk interface bus 2
The maximum number n of IDs that can exist on the system 01 may be uniquely determined at the stage of configuring the system. Only the ID where the disk device was actually stored is the management ID and the managed ID.
The number of the largest ID is assigned to the smallest ID among them as the management ID. Further, for the other IDs in which the disk devices are stored, an ID that is larger than the ID and has the closest number is assigned as a management ID. An example of the system configuration shown in FIG. 1 is shown in the ID management table 130 of FIG. The failure detection LED control means 124 creates the ID management table 130 in this way and holds it in the management information holding means 111.

Next, an LED light emission method for notifying the user which disk device 300 has failed will be described.
When a certain disk device 300 fails, the disk interface control unit control unit 123 sends the disk device 30
A notification is sent if 0 has failed. Based on this information, the RAID control unit 122 queries the failure detection LED control unit 124 about the LED 203 to emit light. The failure detection LED control means 124 determines a management ID associated with the failed disk device 300 based on the ID management table 130.
Is notified to the RAID control unit 122. For the disk device 300 of the management ID, the RAID control unit 122
An LED emission command is issued via the disk interface control unit control unit 123. LED control means 301 of the management disk receiving the LED emission command
Causes the LED 203 corresponding to the managed disk connected to the LED wiring unit 202 via the switch a204 to emit light.

As described above, the management disk emits the LED 203 corresponding to the managed disk, and the user can operate the failed disk device 300 by the emitted LED.
Can be specified. Further, when the failed disk device 300 is removed, the corresponding LED 203 and the switch a 204 are separated from each other.
Recognition by 1 can prevent the LED 203 from emitting light again thereafter.

With the above configuration and operation, according to the present invention, the disk management device is not used, and the RAI
There is an effect that the means for performing D control and the disk case are connected without using a special dedicated line, and the lighting control of the LED indicating the failed disk device can be provided at a low price.

A second embodiment of the present invention will be described with reference to the drawings. The second embodiment monitors the ID of a failed disk device from a nearby disk device, and notifies the host computer that the failed disk device has been removed by the user and that a new disk device has been inserted by the user. It is for notification.

FIG. 4 is a system configuration diagram of a second embodiment of the present invention. Only the differences from FIG. 1 will be described. In the disk case 200, the switches a204-0 to 204-
switches b208-0 to 208- in place of n
n is the connection inspection wiring unit 2 instead of the LED wiring unit 202
09 are provided, but the operation and connection form are the same as those in FIG. Switches b208-0 to 208-
n and the connection investigation wiring unit 209 are collectively referred to as a hot-swap management disk determination unit 210. The hot-swap management disk determination unit 210 includes a failed disk device 301,
The relationship between the management disks that monitor the hot swapping of the disk device 301 is uniquely determined.

Connection checking means 207-0 to 207-n are provided in place of the LEDs 203-0 to 203-n, respectively, and LEs are provided inside the disk devices 300-0 to 300-n.
Instead of the D control means 301-0 to 301-n, connection status recording means 302-0 to 302-n are provided, respectively, and the connection form is the same as that of FIG.

The connection checking means 207-0 to 207-n
The disk devices 300-0 to 300-n depend on whether they are connected to the switches b208-0 to 208-n, respectively.
Is used to check if is stored on the disk case. Connection status recording means 302-0 to 302-
n records that the disk devices 300-0 to 300-n have been inserted and removed.

FIG. 5 is a block diagram of the RAID control program 110 according to the second embodiment. The difference from FIG. 2 is that the failure detection LED control means 124 has been replaced with a hot-swap control means 125. Hot-swap controller 1
Reference numeral 25 is the first like the failure detection LED control means 124.
The ID management table 130 is created by the same method as that of the first embodiment, and held in the management information holding unit 111. The ID management table 130 according to the second embodiment is for specifying a management ID for monitoring the connection checking unit 207 corresponding to the failed disk device 300.

Next, an operation of notifying the host computer 100 that the failed disk device has been removed by the user and that a new disk device has been inserted by the user will be described. The method of creating and holding the ID management table 130 by the hot-swap controller 125 is the same as in the first embodiment.

When a certain disk device 300 fails, a notification of which disk device 300 has failed is sent to the disk interface control means control unit 123 via the disk interface control means 108. Based on this information, the RAID control unit 122 inquires of the hot-line insertion / removal control unit 125 about the connection investigation unit 207 to be monitored. The hot-swap controller 125 includes the ID management table 1
The management ID related to the failed disk device 300 is notified to the RAID control unit 122 based on 30. Management I
D disk device 300, the RAID controller 12
2 is a disk interface control means control unit 123
Issue a connection status acquisition command via

[0042]

According to the present invention, it is not necessary to use a special dedicated line to connect the means for controlling RAID to the disk case, so that the disk case can be managed at low cost. This has the effect.

Furthermore, according to the present invention, there is no need to provide a device dedicated to managing the disk case on the disk interface bus, so that it is possible to manage the disk case at low cost without reducing the number of devices that can be connected to the disk interface bus. There is an effect that can be performed.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram according to a first embodiment.

FIG. 2 is a block diagram of a RAID control program according to the first embodiment.

FIG. 3 is a block diagram of a management information holding unit according to the first embodiment.

FIG. 4 is a system configuration diagram according to a second embodiment.

FIG. 5 is a block diagram of a RAID control program according to a second embodiment.

[Explanation of symbols]

100: Host computer, 101: Host CPU, 102
... Cache, 103, Chipset, 104, Local Bus, 105, System Bus, 106, RAID Control Unit, 107, Bus Interface, 108, Disk Interface Control Unit, 109, Main Memory,
110: RAID control program, 111: management information holding unit, 112: user area, 120: request receiving unit, 1
21 ... request end unit, 122 ... RAID control unit, 123 ...
Disk interface control means control section, 124: failure detection LED control means, 125: hot-swap control means, 1
30: ID management table, 200: disk case,
201: disk interface bus, 202: LE
D wiring section, 203-0 to 203-n ... LED, 204-
0-204-n ... switch a, 205-0 to 205-n
... Bus connection unit, 206 ... LED management disk determination means,
207-0 to 207-n ... connection investigation means, 208-0 to
208-n: switch b, 209: connection investigation wiring section, 2
10 hot-swap management disk determination means, 300-0 to 3
00-n: Disk device, 301-0 to 301-n: L
ED control means, 302-0 to 302-n... Connection status recording means

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Yasunori Kaneda 1099 Ozenji Temple, Aso-ku, Kawasaki City, Kanagawa Prefecture Inside Hitachi, Ltd.System Development Laboratory (72) Inventor Takashi Arakawa 1099 Ozenji Temple, Aso-ku, Kawasaki City, Kanagawa Prefecture (72) Inventor Masahiro Takano 2880 Kozu, Odawara-shi, Kanagawa Prefecture Storage Systems Division, Hitachi, Ltd.

Claims (5)

[Claims] 1. A host computer comprising: a RAID controller and a RAID controller;
A computer system having a D control program, a plurality of disk devices stored in a disk case, and a host computer and a disk case connected by a disk interface bus for transferring data to and from a disk. LED control means for controlling the LED light emission of the disk device is provided. An LED corresponding to each of the plurality of disk devices, a disk device and an LED corresponding thereto are provided on the disk case.
An LED management disk determining means for uniquely determining a disk device that controls ED light emission is provided, and a RAID control program of the host computer includes an ID for associating a certain disk device with a disk device that controls the corresponding LED to emit light. A failure detection LED control means for creating and referencing a management table is provided, and the host computer
A computer system comprising management information holding means for storing a management table. 2. A host computer comprising: a RAID controller and a RAID controller;
A computer system having a D control program, a plurality of disk devices stored in a disk case, and a host computer and a disk case connected by a disk interface bus for transferring data to and from a disk. Connection status recording means for monitoring and recording the connection status between the disk device and the interface bus, and the disk case includes connection investigation means for indicating whether each of the plurality of disk devices is connected to the interface bus. A hot-swap management disk determining means for uniquely determining a disk device and a disk device that monitors whether the disk device is connected to the interface bus is provided. The RAID control program of the host computer includes a disk device and its disk. If the device is A hot-swap controller for creating and referencing an ID management table for associating a disk device for monitoring whether or not it is connected to a host computer is provided, and a management information holding means for storing the ID management table is provided in the host computer. A computer system characterized in that: 3. The computer system according to claim 1, wherein the RAID control program is executed by a host CPU of a host computer. 4. The computer system according to claim 1, wherein said RAID control means has a processor dedicated to RAID control / management, and said RAID control program is a RAID control program.
A computer system which is executed by a processor dedicated to control / management of D. 5. The computer system according to claim 1, wherein the RAID control means and the RAID control program are provided in a RAID device external to the host computer, and the RAID device and the host computer are connected by a data transfer path. A computer system characterized by: