JPH10320236A - Maintenance deterrent device for preventive maintenance - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To detect the occurrence of a failure at an early stage and suppress maintenance by checking status information of a disk control device and operation information of a maintenance application during maintenance. SOLUTION: A disk controller 202 having a duplicated processor and a memory, and a maintenance computer 203 provided in the disk controller for performing maintenance of the disk controller, are provided. Operation information 105 relating to the operation of the installed maintenance application 107 and the disk controller 20
2 and the status information 101, and monitor 102, and as a result of the monitoring,
When it is detected that a failure has occurred in the disk controller, a recovery process 108 for recovering the disk controller to a state immediately before maintenance is performed based on the history 105 of the operation information of the maintenance application.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a maintenance operation of a control device including a plurality of processors and memories operated by a program.

[0002]

2. Description of the Related Art When a control device composed of a plurality of processors and a memory operated by a program is maintained by a dedicated maintenance machine (hereinafter abbreviated as "service processor SVP") provided in the device, the service processor is, for example, a personal computer. The following is a description of a technique implemented by a computer application referred to as “computer”.

[0003] When the control device is used in a non-stop state in conjunction with the host machine, the maintenance work must be performed under the use state, and errors in the maintenance operation are not allowed. For this reason, immediately before the maintenance of the control device is performed, status information is transmitted from the control device (for example, a state in which the channel adapter of the disk control device is closed for maintenance).
And checks whether the system maintenance, data loss and performance degradation of the control device are not caused by the continuation of the maintenance. Further, when the maintenance is interrupted, a message is displayed to alert the maintenance staff of an instruction or a caution.

[0004] Monitoring and detection of faults in the control device,
Techniques for failure analysis and recovery are already known. For example, a fault monitoring technique is disclosed in Japanese Patent Application Laid-Open No. 5-324377, entitled "Processor Monitoring System". The recovery from a failure is described in Japanese Patent Application Laid-Open No. 6-282451, entitled "Method of Restoring System Down of Device with Microprocessor".

[0005] The feature of the restoration technique for these faults is that they are automatically performed.

[0006]

However, the state of the control device is constantly shifting. However, in the above-described conventional technology, a check is not performed in accordance with the transition of the device state. That is, the failure of the control device during the maintenance is not recognized. For this reason, if a failure occurs during maintenance, the failure may cause secondary failures such as system down, data loss, and deterioration of control device performance due to the maintenance being performed.

[0007] For example, in the above-mentioned known example, it is possible to automatically perform from failure monitoring to recovery. However, during the recovery process, when a maintenance worker performs another maintenance operation, a secondary failure occurs. There is a possibility. It is also possible to suppress the operation of maintenance personnel during recovery, but there are serious and trivial failures, and it is difficult to judge the failure, and there is no conventional technology to solve this.

SUMMARY OF THE INVENTION An object of the present invention is to prevent a system failure, data loss, and performance degradation of a control device due to maintenance work, and how the control device checks and suppresses the status of the control device during maintenance. It is an issue.

[0009]

Means for Solving the Problems In order to solve the above problems, the present invention mainly employs the following constitution.

A disk controller having a duplicated processor and a memory, a duplicated disk drive, and a maintenance device provided in the disk controller for performing maintenance of the disk controller and the disk drive. A computer, wherein the maintenance computer monitors operation information relating to operation of a maintenance application mounted on the maintenance computer, and status information of the disk control device and the disk drive device. As a result, when it is detected that a failure has occurred in the disk control device or the disk drive device, a maintenance suppression device for forcibly terminating the corresponding maintenance application being operated and automatically interrupting maintenance. .

A disk controller having a duplicated processor and a memory, and a maintenance computer provided in the disk controller for performing maintenance of the disk controller are provided. Monitoring operation information relating to the operation of the maintenance application installed in the maintenance computer and status information of the disk controller, and as a result of the monitoring, it is detected that a failure has occurred in the disk controller. And performing a recovery process for recovering the disk control device to a state immediately before maintenance based on a history of operation information of the maintenance application.

[0012]

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

FIGS. 1 and 2 show an embodiment of the present invention. In this case, 101 is a function for acquiring the status information of the disk control device, 102 is a function for performing a mutual check between the operation information and the disk control device status information, 103 is a table showing the correspondence between the operation information, the disk control device and the application to be handled, 104, an application interface; 105, a table for storing operation history; 106, a task for exchanging a disk controller with an application I / F; 107, a maintenance application;
08 is a recovery application, 109 is a message display application, 120 is a CRT, 12
1 is a keyboard, 122 is a mouse, 123 is a floppy disk drive, 201 is a host machine, 202 is a disk controller, 203 is a service processor, 204
Is a common bus, 205 and 206 are channel adapters, 20
7, 208 are disk adapters, 209 is a non-volatile cache memory, 210 is an internal bus, 220 is a disk drive, 221 is a hard disk unit, 222 is a power supply for supplying power to the disk controller and drive, and 223 is a SCSI. Cable, respectively.

FIG. 2 shows a disk subsystem device.
The disk subsystem device includes a disk control device 202
(Hereinafter abbreviated as “DKC”) and the disk drive 2
20 (hereinafter abbreviated as “DKU”).

The DKC 202 has channel adapters 205 and 206 (hereinafter referred to as “C”) for controlling information between the hosts 201.
HA ”. ), Disk adapters 207 and 208 (hereinafter abbreviated as "DKA") for controlling information between DKUs 220, and a non-volatile cache memory 209 (hereinafter abbreviated as "CM") for speeding up disk access. It is configured.

These are all duplicated, and are connected by the internal bus 210 to enable communication. C
The HAs 205 and 206 and the DKAs 207 and 208 are equipped with a microprocessor (hereinafter abbreviated as “MP”). The DKC 202 is equipped with a service processor SVP 203 for maintenance.
205, 206 / DKA 207, 208 / CM 209 are connected by a common bus 204 to enable communication.

The DKU 220 has a disk group 221 (hereinafter abbreviated as "HDU") for storing customer information, and
A power supply 222 for supplying power to the DKU and DKC is mounted. The HDU 221 and the DKAs 207 and 208 are connected by a SCSI cable 223.

FIG. 1 shows the configuration of an embodiment of the present invention. The left part surrounded by the dotted line in FIG.
3 shows the internal configuration, and the left part thereof shows the disk control device 202.

The maintenance worker starts the application AP 107 by using the keyboard 121 or the mouse 122 to execute maintenance. After that, the AP 107 divides the input specific maintenance procedure and issues a command unit (a command unit such as maintenance blockage, replacement, or recovery) to the API I / F 104. AP I / F10
4 stores the received command unit in the history 105 and D
It is issued to the KC status information acquisition 101. That is, a trigger is issued so that DKC status information can be acquired for each command unit.

In the DKC status information acquisition, the current DKC status is acquired from the DKC communication 106, and the operation & status (operation information and device status) mutual check 102 (current information is
(Triggering whether or not the condition expression shown in FIG. 3 is satisfied) is started (triggered).

The operation & status mutual check 102 acquires operation information from the history 105 via the API / F 104. Also, a conditional expression 103 is acquired as a check condition,
Based on the current DKC status information, operation information, and conditional expression, the 102 performs a mutual check, and issues an instruction for the next process to the AP.
Perform to I / F104.

The API / F 104 issues a DK
It issues a command to the C communication 106, issues a command to the recovery AP, issues a command to the message display AP 109, and issues a command to the AP 107. In the case of issuing a command to the DKC communication 106, the DKC communication 106 that has received the command
Issue commands to 05-208.

As described above, in the present invention, the maintenance of the control device is performed by issuing a command from the service processor mounted on the control device to each of the microprocessors. Check the status of the control device. By executing this check, a condition expression for acquiring the state information of the apparatus as needed and determining the next processing based on the operation information and the apparatus state information is required. Using these, the command issued from the maintenance application is distributed to each task by the application I / F. By doing so, it is possible to check the status of the control device in units of command issuance, and to prevent a system down, data loss, and performance degradation of the control device.

FIG. 3 shows a conditional expression, which means that an inconvenient state is tabulated in advance, processing is performed in accordance with the table, and maintenance work can be continued when the condition does not apply to the table. .

The conditional expression is a correspondence table between the operation information and the DKC state, and the next processing is determined based on the two states. This conditional expression is roughly divided into three items: operation information, DKC device state, and coping AP.

The operation information includes the name of the application (AP) that has performed the maintenance (the
), The type of maintenance, the target maintenance part, the detailed part of the target maintenance, and the state of the target part. Here, for example, CHA1 which is a package board
Is replaced, the CHA1 is duplicated, so that the board to be replaced is placed in a maintenance closed state for maintenance such as replacement, and replacement is attempted.

The device status of the DKC describes a portion of another resource, a detailed portion of another resource, and a status of another resource portion.
Here, for example, if the status of one board of the CHA 1, that is, the status of other resources is not grasped, the whole system may be down even if it is duplexed.

The target AP at the time of recovery or message display is described in the coping AP. In this case, for example, if the system is going to go down, the recovery AP 1 is operated to recover the board whose maintenance is blocked.

[Example] shown in the lower part of FIG. 3 is that the operation information is "first exchange of CHA by exchange AP". In addition, the DKC state is “the third DKA is a fault blockage”. As a result, the "AP for displaying messages"
Is started and a warning is displayed to the maintenance staff by a message. As described above, the conditional expression is composed of the operation information, the device state,
Each item in the coping AP is specified, and many possible conditional expressions are completed. In other words, the condition expression includes all possible situations and how to deal with them (such as displaying a message or operating an appropriate recovery AP).

FIG. 4 shows a flowchart of the API / F 104.

The API / F 104 is roughly composed of three commands: a command 401 from the maintenance AP, a command 402 from the DKC communication, and a command 403 from the mutual check.

First, when there is a command 401 from the maintenance AP, the operation history of the command is stored 410. After that, if it is received 411 from the maintenance AP,
A command 413 is issued to the DKC communication. In the case of the reception 412 from the recovery AP, the response destination is determined 414, and in the case of the response to the DKC communication, a command is issued 415 to the DKC communication.

Next, when there is a command 402 from the DKC communication, the operation history of the command is stored 420. After that, if it is received 421 from the maintenance AP,
A command 423 is issued to the DKC communication. In the case of the reception 422 from the recovery AP, the recovery A
Issue command 424 to P.

Next, the command 403 from the mutual check
If there is, the operation history of the command is stored 430. Thereafter, it is determined 431 whether to suppress the operation. If the operation is to be suppressed, the target recovery AP is searched 433 and a command is issued 434 to the target recovery AP.

If not suppressed, the target maintenance AP
A normal report 432 is made.

The three checks are looped to constantly monitor.

FIG. 5 shows a flowchart of the DKC status information acquisition 101. In the DKC status information acquisition 101, the AP
When there is a command 501 from the I / F, the DKC status information is acquired 511 from the DKC, and an instruction is issued 511 to the operation & DKC status mutual check. By looping this process, the latest DKC state information is always obtained.

FIG. 6 shows a flowchart of the operation & status mutual check 102. Mutual check of operation & status 10
In 2, when the mutual check execution instruction 601 is given,
The target conditional expression is searched and acquired 611. Thereafter, a check 612 of the conditional expression, the DKC state and the operation history is performed.
As a result 613, if maintenance is to be suppressed, API / F
Then, an instruction 615 for recovery processing is issued. If the maintenance is not suppressed, a normal report is issued to the AP I / F.
14 By looping this process, it is possible to always perform the operation and state mutual check.

In the above description, the information on the device state is
Although the status of the disk control device has been mainly targeted, the status of the disk drive device can also be included in addition to this.

[0040]

According to the present invention, the control device is required to be maintained in a non-stop state, and a secondary failure due to the maintenance can be avoided in advance and during the maintenance. Further, there is an effect that maintainability and reliability are improved by improving operability by displaying a message and automating recovery processing.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a first embodiment of the present invention.

FIG. 2 is a hardware configuration diagram for implementing an embodiment of the present invention, and particularly a hardware schematic diagram of a magnetic disk device.

FIG. 3 is a diagram illustrating a table indicating correspondence between operation information, a control device, and a countermeasure application;

FIG. 4 is a flowchart for implementing an embodiment of the present invention.

FIG. 5 is a flowchart for realizing an embodiment of the present invention.

FIG. 6 is a flowchart for implementing an embodiment of the present invention.

[Explanation of symbols]

101: Function for acquiring status information of a disk controller 102: Function for performing mutual check between operation information and disk controller status information 103: Table showing correspondence between operation information, disk controller and corresponding application 104: Interface of application 105 : Table for storing operation history 106: task for exchanging disk controller and application I / F 107: application for maintenance 108: application for recovery 109: application for message display 120: CRT 121: keyboard 122: Mouse 123: Floppy disk drive 201: Host machine 202: Disk controller 203: Service processor 204: Common bus 205, 206: Channel adapter 207, 08: Disk Adapter 209: non-volatile cache memory 210: internal bus 220: a disk drive 221: Hard Disk Unit 222: power supply for providing power to the disk controller and the drive unit 223: SCSI cable

Claims (3)

[Claims] A disk controller having a duplicated processor and a memory; a duplicated disk drive; and a disk controller provided for performing maintenance of the disk controller and the disk drive. A maintenance computer, wherein the maintenance computer monitors operation information related to operation of a maintenance application mounted on the maintenance computer, and status information of the disk control device and the disk drive device, As a result of the monitoring, when it is detected that a failure has occurred in the disk control device or the disk drive device,
A maintenance suppression device for forcibly terminating a corresponding maintenance application being operated and automatically suspending maintenance. 2. A disk control device having a duplicated processor and a memory, and a maintenance computer provided in the disk control device for performing maintenance of the disk control device, wherein the maintenance computer is Monitoring operation information related to the operation of the maintenance application mounted on the maintenance computer and status information of the disk controller, and as a result of the monitoring, it is detected that a failure has occurred in the disk controller. And performing a recovery process for recovering the disk control device to a state immediately before maintenance based on a history of operation information of the maintenance application. 3. A disk controller having a duplicated processor and a memory, and a maintenance computer provided in the disk controller to perform maintenance of the disk controller, wherein the maintenance computer is Monitoring the operation information on the operation of the maintenance application mounted on the maintenance computer and the status information of the disk control device, and monitors the operation information, the status information, and the failure handling application. Various failure handling tables comprising a combination are provided in advance, and the operation information and the status information are monitored and compared with the information and the table. A maintenance suppressing device for performing a recovery process for recovering the device.