JPH05265663A - Diagnostic system for magnetic disk - Google Patents

Abstract

PURPOSE:To improve the diagnostic efficiency of a magnetic disk device by stacking the advancing status of diagnosis, automatically continuing the diagnosis at the time of turning on a power supply after turning off it, or executing automatic repairment at the time of detecting a temporary fault, and at the time of detecting a fault, automatically reporting the detection to a maintenance center to quickly substitute the failed magnetic disk device. CONSTITUTION:The diagnosis of magnetic disk devices 9, 10 is always executed. The diagnosis is executed with a microprocessor 4 and the advancing status of the diagnosis and fault detecting status are stored in a hard disk 7, displayed on a panel 6 and reported to the maintenance center through a modem 8. Rediagnosis or the restart of diagnosis is commanded from the maintenance center and the panel 6 to the microprocessor 4 through a service processor 5. Since the diagnosis of the magnetic disk device 9, 10 can be automatically started even when the power supply is turned on after turning off it, the diagnosis can be efficiently executed, and at the time of detecting a fault, automatic repairment can be attained, attaining the quick detection and repairment of the fault.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a diagnostic system for a peripheral storage device which is a peripheral device of a computer.

[0002]

2. Description of the Related Art A conventional method is disclosed in Japanese Patent Laid-Open No. 2-216542.
As described in 1), the operating rate is monitored in advance, and when the operating rate becomes low, the system is disconnected from the processing device to diagnose the system, and continuous diagnosis when the power is turned on and off is not considered.

[0003]

According to the above-mentioned conventional method, since the progress status of the diagnosis is not stuck, the diagnosis cannot be restarted by turning on the power after turning off the power. As the number of processing devices increases and the diagnosis time increases, the power cannot be turned off during the process, so the maintenance staff cannot be interrupted.

An object of the present invention is to read a disk medium.
By stacking the progress status of the diagnosis and the failure detection status, it is possible to suspend and restart the diagnosis when the power is turned off and on, and the restart is automatically performed, so that the burden on the maintenance staff for monitoring can be reduced.

Further, since READ diagnosis is always performed, when a failure is detected during diagnosis or a temporary failure is detected, an automatic notification is given to the maintenance center so that the disk medium can be replaced promptly. In the case of a mild disorder, it is repaired automatically because it is repaired early. In addition, automatic repair will be performed according to instructions from the maintenance center. In addition, the progress display and error display on the service processor for the purpose of READ diagnosis on the site and re-diagnosis when an error is detected. In addition, the diagnostic condition is changed to diagnose a specific failure site.

[0006]

In order to achieve the above object, the following method is adopted.

(1) READ diagnosis of the disk medium is continuously executed. The progress status at the time of executing the diagnosis is always stacked in the disk control device, and the READ diagnosis can be interrupted and automatically restarted even if the power supply of the disk control device is turned off and on.

(2) During READ diagnosis, the type of failure, the number of times of detection, and the address are stacked, and when a certain threshold is exceeded, the maintenance center is automatically notified.

(3) Since the READ diagnosis is always executed,
The diagnostic progress status at that time is stacked on the memory or the hard disk and displayed on the panel via the service processor. In addition, the failure location is identified by displaying the failure detection location on the disk medium on the service processor,
Rediagnosis can be performed by changing the conditions (slice level change, READ frequency change, etc.).

(4) When an abnormality of the disk medium is detected by the READ diagnosis, the service processor stacks the type and location of the failure and notifies the maintenance center so that the service processor automatically repairs it according to an instruction from the maintenance center ( Alternate track processing) to the DKC. In the case of a temporary failure (correctable error), the service processor causes the DKC to automatically perform repair.

[0011]

According to the present invention, the execution result, the progress status, and the failure detection portion at the time of READ diagnosis of the disk medium are stored in the memory of the DKC and the service processor, and further stacked on the hard disk. From service processor to D
Since the READ diagnosis can be instructed to the KC, the READ diagnosis can be automatically restarted when the power is turned on and then on.

The service processor automatically notifies the maintenance center when a failure is detected via the communication line.
Instructing rediagnosis and automatic repair of the disk medium, the service processor receives the instruction and instructs the DKC to perform rediagnosis and automatic repair.

[0013]

1 is a block diagram of a magnetic disk subsystem according to an embodiment of the present invention. The magnetic disk control device (3) has therein a microprocessor (4), a service processor (5), a panel (6) as a display device, a hard disk (7) as a storage device, and a modem (8) as a communication control device. is doing. A host computer (1) and a channel device (2) are connected to the upper layer, and magnetic disk devices (9) and (10) are connected to the lower layer.

A command for READ / WRITE of (9) and (10) is issued from (1) to (3) via (2). In (4), always (9), (10)
READ diagnosis is executed. However, (9) or (10)
When (1) is in use, the READ diagnosis is executed when the frequency is low, but when it is not low, the READ diagnosis is suspended.
The progress status and failure detection status of the READ diagnosis in (9) and (10) are transferred from (4) to (5) and accumulated in (7). Further, it is displayed in (6) for reporting to maintenance personnel, or displayed at a remote maintenance center via (8).
In (5), the failure of (9) and (10) is automatically notified to the maintenance center, and further from the maintenance center or (6),
Rediagnosis of (9) and (10) or failure repair can be instructed via (5) (4).

The interruption / restart of the READ diagnosis is instructed to (4) by (5) accumulating the progress status in (7).

[0016]

According to the present invention, the magnetic disk medium can be continuously diagnosed even if the power of the disk controller is turned off and on.

It is possible to notify the maintenance center automatically when a temporary failure is detected, which can be saved by retry by diagnosis, or when a failure is detected.

Re-diagnosis can be performed by specifying the faulty part and changing the diagnostic conditions.

The magnetic disk can be automatically repaired when a temporary failure or failure is detected.

[Brief description of drawings]

FIG. 1 is a block diagram of a magnetic disk subsystem according to an embodiment of the present invention.

[Explanation of symbols]

 3 ... Magnetic disk control device, 4 ... Microprocessor, 5 ... Service processor, 6 ... Panel, 7 ... Hard disk, 9, 10 ... Magnetic disk device.

Claims (4)

[Claims] 1. A READ diagnosis is executed regardless of online / offline for the purpose of preventive maintenance of a disk medium.
Stack the progress of READ diagnosis and failure detection,
Even if the magnetic disk controller is turned off and then turned on, R
A magnetic disk drive diagnostic method characterized in that EAD diagnosis can be automatically restarted from the position when the power was turned off. 2. A magnetic disk device diagnosis characterized by counting the number of occurrences of temporary failures saved by retries during READ diagnosis of a disk medium, and automatically reporting to a maintenance center if a certain threshold is exceeded. method. 3. While the READ diagnosis of the disk medium is being executed, the progress status of the diagnosis and the failure detection status are displayed on the panel of the service processor or the remote maintenance center, and the failure occurrence site is specified and the diagnosis condition is changed. Re-diagnosis (instructions can be made from the panel or a remote maintenance center) and the magnetic disk drive diagnostic method. 4. When a disk medium abnormality is detected during execution of READ diagnosis, the disk medium abnormality is automatically repaired, and this restoration is automatically performed when the abnormality is detected and automatically repaired by an instruction from a remote maintenance center. A magnetic disk drive diagnostic method characterized by carrying out.