JP2001202202A - Disk controller and recording medium with recorded program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To protect the data of a cache memory at the time of resetting a computer in a disk controller for writing data from a computer in a cache memory, and then for writing the data in a disk. SOLUTION: Normally, a controller 1(131) temporarily preserves data from a computer 1(101) in a cache memory 15, and then writes the data in a disk 16. When the computer 1(101) is restarted by any factor, and reset is notified to the controller 1(131), the notification is notified to an emergency data processing part 182 of a controller 2(132) by a reset recognition notifying part 171. Then, the data received by the controller 1(131) are added to the cache memory 15, or the data are written in the disk 16 before data to be processed by the controller 2(132) so that the data can be prevented from being lost.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer management system which manages a plurality of computers having a shared disk and takes over the processing to another computer when one computer fails. The present invention relates to a disk control device for controlling a shared disk and a recording medium on which a program used in the disk control device is recorded.

[0002]

2. Description of the Related Art FIG. 4 is a block diagram showing a configuration of a computer management system using a conventional disk controller. 4, a plurality (two in this case) of computers 1 (101) and computers 2 (102) are managed by the disk controller 11.

[0003] The disk control device 11 is a computer 1
It has a controller 1 (131) communicating with (101) via the input / output unit 121, and a controller 2 (132) communicating with the computer 2 (102) via the input / output unit 122. Controller 1 (131) and Controller 2
(132) is connected by a bus 14, to which a cache memory 15 and a disk 16 are connected.

Next, the operation will be described. Now, assuming that the computer 1 (101) is operating, and if the data processed by the computer 1 (101) is to be saved, as shown in FIG.
Receives the data, temporarily stores it in the cache memory 15, and then saves it on the disk 16 as shown in FIG.

[0005] Here, the computer 1 (101) is HW
When a restart is caused due to a failure, an OS panic, or any other factor, a reset is transmitted to the disk control device 11 as shown in FIG. The disk control device 11 recognizes the reset from the computer 1 (101) and performs an initialization operation of the cache memory 15 as shown by the symbol in FIG. This initialization is performed on the computer 1 (10
This is an operation necessary for recognizing and configuring resources connected to the computer 1 (101) at the time of startup of 1).

After the initialization of the cache memory 15, the processing of the computer 1 (101) is performed by the computer 2 (102).
The data is written to the cache memory 15 and the disk 16 by the controller 2 (132).

As a management system for a plurality of computers, there is another system described in, for example, JP-A-10207855. This publication discloses a computer provided with a disk cache memory.

[0008]

However, when the above reset occurs, if the normal data sent to the disk controller 11 before the reset still exists in the cache memory 15, Since the initialization operation is performed before the data is written to the disk 16, the data in the cache memory 15 is erased and there is a problem that important data is lost. Further, the above-mentioned Japanese Patent Application Laid-Open No. 10207855 also has a similar problem that the contents of the disk cache memory become invalid at the time of reset.

The present invention has been made to solve the above problem, and has as its object to prevent data in a cache memory from being lost at the time of reset.

[0010]

In order to achieve the above object, a disk control apparatus according to the present invention comprises a computer (for example, the computer 1 (10) in the embodiment).
1)) a cache memory (the same cache memory 15) for temporarily holding the data, and a plurality of computers, each of which is provided with a corresponding one of the computers and receives data from one computer and writes the data into the cache memory. A plurality of control means (the input / output unit 12) for writing data in the cache memory to the disk (the disk 16)
1, 122, Controller 1 (131), Controller 2
(132)) and a plurality of reset recognition notifying means (the input / output units 121, 122, and 122) provided respectively corresponding to the plurality of control means and receiving the reset notification from the one computer and notifying the reset notification to the other. Reset recognition notification unit 17
1 and 172) and a plurality of emergency data processing means (the emergency data processing units provided respectively corresponding to the plurality of control means and protecting the data in the cache memory in response to the reset notification from the reset recognition notification means). 181,
182).

A recording medium on which a program according to the present invention is recorded receives data from a computer, writes the data in a cache memory, controls the writing of the data in the cache memory to a disk, and issues a reset notification from the computer. A program for executing a receiving process to be received and an emergency data process for protecting data in the cache memory in response to the received reset notification are recorded.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a computer management system using a disk control device according to the present invention. Parts corresponding to those in FIG. 4 are denoted by the same reference numerals, and redundant description is omitted.

In this embodiment, the computer 1 (10
1) In the disk controller 11 for receiving and storing data from the computer 2 (102), the controller 1 (13) in the disk controller 11 as shown in FIG.
1), controller 2 (132), computer 1 (1
01), and reset recognition notification units 171 and 172 for performing communication upon receiving a reset from the computer 2 (102). Such communication may be communication by HW signals, command communication on the bus 14 connecting the controllers, or the like.

Further, the controller having received the notification,
Emergency data processing units 181 and 182 are provided, which determine whether there is any data being processed by the controller that has issued the notification and, if there is data, fetch the data into its own processing.

With the above arrangement, when one controller detects a reset from the computer, the other controller replaces the writing of data received by the one controller, thereby preventing data loss. I have to.

Next, the operation will be described. First, a case where the computer 1 (101) is activated after the power is turned on will be considered. At this time, a reset is notified from the computer 1 (101). This reset is performed by the reset recognition notifying unit 17.
The other computer 2 (102) is notified by 1 and 172. However, since the power of the computer 1 (101) has been turned on, there is no data received from the computer 1 (101) before that, and the computer 2 (102) has no data. No data is processed in 102). Therefore, the operation is the same as before.
There is no problem even if the method according to the present embodiment is realized.

During normal operation, the computer 1 (10
The data received from 1) is temporarily stored in the cache memory 1 by the controller 1 (131) as shown in FIG.
5 and then as shown in FIG.
6 is written.

Next, during the normal operation, the computer 1
The case where (101) is restarted for some reason and a reset is notified to the controller 1 (131) as shown in FIG. 3 will be described. Reset is controller 1
When the notification is sent to (131), the notification is sent to the controller 2 (1) by the reset recognition notifying unit 171 as shown in FIG.
It is notified to the emergency data processing unit 182 of 32). In response to this, the emergency data processing unit 182 operates as shown in FIG. 3 and the controller 1 (131) as shown in FIG.
Adds the data received from the computer 1 (101) to the cache memory 15 as data to be processed by itself, or writes the data of the controller 1 (131) to the disk 16 before the data to be processed by the controller 2 (132). Perform processing.

According to the present embodiment, the reset recognition notifying units 171 and 172 for notifying the other controllers that the reset has been detected, and the data which should have been performed by the controller which has notified when notified, are transmitted to the self controller. By providing the emergency data processing units 181 and 182 for protecting data by being incorporated in the processing of the above, writing of data from the cache memory 15 to the disk 16 which is not originally processed by the reset is executed by being replaced by another controller. Therefore, it is possible to prevent data from being lost.

A memory such as a ROM and a RAM for storing the program of the CPU in the computer system constituting the disk control device 11 constitutes a recording medium on which the program according to the present invention is recorded. This recording medium stores a program for executing the operation of the above-described embodiment. Further, as such a recording medium, a semiconductor storage device, various disk media, a magnetic recording medium, or the like can be used.

[0021]

According to the present invention, a plurality of reset recognizing means for notifying a control means such as another controller that a reset from a computer has been detected, and a control means for notifying when a reset has been notified. By embedding the data that was supposed to be done in its own processing, by providing multiple emergency data processing means to protect the data,
Since writing of data from the cache memory to the disk, which is not originally processed by the reset, is executed by being replaced by another control means, it is possible to prevent loss of data in the cache memory.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a computer management system using a disk control device according to an embodiment of the present invention.

FIG. 2 is a block diagram showing an operation of the computer management system according to the embodiment of the present invention.

FIG. 3 is a block diagram showing an operation of the computer management system according to the embodiment of the present invention.

FIG. 4 is a block diagram showing a configuration of a computer management system using a conventional disk control device.

FIG. 5 is a block diagram showing the operation of a conventional computer management system.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 101, 102 Computer 11 Disk controller 121, 122 Input / output part 131, 132 Controller 14 Bus 15 Cache memory 16 Disk 171, 172 Reset recognition notification part 181, 182 Emergency data processing part

Claims (6)

[Claims] A cache memory for temporarily holding data from a computer; a cache memory provided for each of a plurality of computers; receiving data from one computer and writing the data in the cache memory; A plurality of control means for writing data to a disk; and a plurality of control means respectively provided corresponding to the plurality of control means.
A plurality of reset recognition notifying means for receiving a reset notification from one computer and notifying the reset notification to the other; and the cache memory provided respectively corresponding to the plurality of control means and in response to the reset notification from the reset recognition notifying means. A plurality of emergency data processing means for protecting said data. 2. The data protection device according to claim 1, wherein the data received by the control means corresponding to the one computer is written to the cache memory by the control means different from the control means and the emergency data processing means corresponding thereto. The disk control device according to claim 1, wherein: 3. The protection of the data by writing the data received by the control means corresponding to the one computer and the emergency data processing means corresponding to the control means different from the control means to the disk. The disk control device according to claim 1, wherein: 4. A control process for receiving data from a computer and writing the data in the cache memory and writing the data in the cache memory to a disk, a receiving process for receiving a reset notification from the computer, and the received reset notification. And a program for executing an emergency data process for protecting data in the cache memory according to the above. 5. The recording medium according to claim 4, wherein said emergency data processing protects said data by writing data received from said computer into said cache memory. 6. The emergency data processing according to claim 4, wherein said data is protected by writing data received from said computer to said disk.
A recording medium on which the program described above is recorded.