JPH0713704A - Magnetic disk controller - Google Patents

Abstract

PURPOSE:To easily detect the destruction of data when the 1st and 2nd magnetic disks are possibly destroyed by the fault factors such as the thunderbolt, etc., by writing the data written in the 1st magnetic disk into the 2nd disk with logical inversion. CONSTITUTION:A magnetic disk controller 1 writes the data that are transferred to an input/output control circuit 2 from a host device 8 by a writing command into a magnetic disk 4 via an input/output circuit 3. Meanwhile the data which are logically inverted by an input/output inverting circuit 5 are written into a magnetic disk 6. When a reading command is given from the device 8, the data are read out of both disks 4 and 6. The data read out of the disk 4 are sent to the device 8, and at the same time the data of both disks 4 and 6 are compared with each other by a comparing/deciding circuit 7. Then the logically discordant data are stored in an error register included in the circuit 7.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a magnetic disk controller for checking whether data written in a magnetic disk device is correct.

[0002]

2. Description of the Related Art Conventionally, a technique has been used in which two sets of information recording means such as magnetic disks are provided and information is double-written to improve the reliability of information recording. As a method to check whether the data written on each magnetic disk in this technology is correctly recorded,
The same data is written in two storage devices, the same data written in the first storage device and the second storage device is read, and both data are compared to detect an error. Checks are being made by multiplexing. This type of technique is disclosed in, for example, Japanese Patent Laid-Open No. 4-20902.
No. 2 publication.

[0003]

However, even if the same data is written in the two storage devices as described above, the data at the same address in the first and second storage devices are the same due to a failure factor such as lightning strike. When the contents are destroyed and the data in the two storage devices after the destruction are the same, there is a problem in that the destruction of the data cannot be detected. That is, when the data “1001” is written to the first and second magnetic disks, the data “1001” becomes “1000” due to a failure factor, and the same “1000” is written to both disks. Cannot detect data corruption. As one of the measures against this problem, there is a method of providing more storage devices, storing the same data in each storage device, reading the written data, taking a majority decision of the data, and detecting an error. However, there is a problem that the cost increases because more connecting means, storage devices, etc. are required. In view of such a problem, the present invention realizes a device that easily finds corrupted data from data stored in a storage device and improves the reliability of the data.

[0004]

[Means for Solving the Problems]

The present invention relates to a magnetic disk controller for connecting a host device and first and second magnetic disks so as to input and output data to and from each other. In the first magnetic disk, the data sent from the host device remains unchanged. Writing means, means for logically inverting and writing the data sent from the host device to the second magnetic disk, data written to the first magnetic disk and the data written to the second magnetic disk. This is achieved by including means for reading the logically-inverted data and comparison / determination means for comparing the data of both systems and determining the logically inconsistent data.

[0005]

The data sent from the host device is written to the first magnetic disk as it is, and the data is logically inverted and written to the second magnetic disk to write the respective data of the two magnetic disks. At the same time, by comparing the data of both read systems, it is possible to check the data in which a logical mismatch is found.

[0006]

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

FIG. 1 is a block diagram showing an embodiment of the present invention. The magnetic disk control device 1 in FIG. 1 includes an input / output control circuit 2 for controlling the input / output of commands and data from a higher-level device 8, an input / output circuit 3 for inputting / outputting data to / from a magnetic disk 4, and data inverted to a magnetic disk 6. Input / output inversion circuit 5 for inputting / outputting the magnetic field, magnetic disk 4 and magnetic disk 6
It is composed of a comparison / determination circuit 7 for comparing the data of both systems written respectively in and to check for an error.

Next, the operation of the magnetic disk controller 1 in this configuration will be described. When a write command is issued from the higher-level device 8 to the magnetic disk control device 1 and data (for example, binary data of “1”) is sent, the input / output control circuit 2 sends to the input / output circuit 3 and the input / output inverting circuit 6, respectively. Send the same data. The data sent to the input / output circuit 3 is written to the magnetic disk 4 as it is (for example, “1”) by the write driver 31.
The data sent to the input / output inversion circuit 6 is logically inverted ("1" is inverted to "0") by the inversion write driver 51 and written to the magnetic disk 6.

When a read command is issued from the host device 8, the read driver 32 is used to read the data written on the magnetic disk 4 and send it to the comparison / determination circuit 7 and the input / output circuit 2. The input / output circuit 2 sends the data as a read result to the higher-level device 8. On the other hand, in the read inversion driver 52, at the same time as the read by the read driver 32, the logic inversion data written in the magnetic disk 6 is read out and logically inverted again to restore the original data which is not logically inverted, and the comparison / determination circuit 7 Send to. The comparison / determination circuit 7 detects the error by comparing the data stored in the magnetic disk 4 and the data stored in the magnetic disk 6 read by the read driver 32 and the read / reverse driver 52, respectively. When a data error is detected, the address value of the erroneous data is transmitted to the host device 8. In this way, it becomes possible to easily detect a data write error.

As shown in FIG. 2, when writing the data "1001" to the first magnetic disk and the inverted data "0110" to the second magnetic disk, the external data is written. It is assumed that erroneous data of "1000" is written in the first magnetic disk due to the failure from. When reading the data, "100" which is the reverse of "0110" stored in the second magnetic disk is re-inverted.
1 "is compared with the data" 1000 "of the first magnetic disk. In this way, a data write error can be easily detected, and the address value at which the error occurred is sent to the host device. , The location of data loss can be identified.

In this embodiment, when the data is read from the magnetic disk 6 in which the inverted data is written, the data is read from the read inversion driver 52, that is, the magnetic disk, and logically inverted again to restore the original data. Although a driver is used, a configuration using a read driver that does not perform inversion is also possible. In that case, the comparison and determination circuit 7 performs an operation of comparing the logically inverted data with the data that is not logically inverted.

FIG. 3 is an example of an operation flow in the control circuit of the present invention.

A number of processing commands are issued from the higher-level device 8 to the magnetic disk control device 1 (30
1). If it is a write command, the data transferred from the higher-level device 8 is written to the magnetic disk 4 as it is, and the logically inverted data is written to the magnetic disk 6 (302).

When a read command is issued from the host device 8, the data is read from the magnetic disks 4 and 6.
The data read from the magnetic disk 4 is transferred to the host device through the input / output control circuit 2 as an output result, and is also sent to the comparison / determination circuit 7. Magnetic disk 4
The logical inversion data written in the magnetic disk 6 at the same time as the data reading in the above is read from the magnetic disk 6 by the input / output inversion circuit 5 and is logically inverted again to be returned to the original data which is not logically inverted. It is sent to the comparison / determination circuit 7 (303).

In the comparison / determination circuit 7, the magnetic disks 4, 6
The data read from are compared for logical matching (303). If the data match, it is judged that there is no error (304), but if the data does not match, it is judged that some error occurred in writing the data, and the address value of the data is stored in the error register (comparison). It is stored in the determination circuit 7 (not shown, the same applies hereinafter) (305).

When it is necessary for the host device 8 to confirm whether or not the data written on the magnetic disk is abnormal, by issuing an error data processing command, the input / output control circuit 2 causes the comparison / determination circuit to judge. The address value stored in the error register 7 is transferred to the higher-level device 8 to initialize the error register (306).

In the above configuration, the address value of the write error is stored in the error register of the comparison / determination circuit 7,
The error register has a sufficient capacity to store a plurality of address values so that when an error data processing command is instructed from the higher-level device 8, the address value that is in error for the first time is sent to the higher-level device 8. However, in the operation flow shown in FIG. 4, when an error occurs in the match determination 404, the address value of the errored data is sent to the higher-level device 8 each time, and the error location may be reported. Is.

[0018]

EFFECTS OF THE INVENTION A data error which is difficult to detect by a conventional method of checking the normality of data written on a magnetic disk by comparing data of both systems using two magnetic disks, for example, Even when data of the same portion written in two magnetic disks are destroyed so as to match due to a failure factor such as lightning strike, by using the present invention, it is possible to easily and effectively without increasing the storage device for comparison. It is possible to detect the destruction of the data written on the magnetic disk.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

FIG. 2 is a diagram showing a data storage state according to an embodiment of the present invention.

FIG. 3 is an operation flowchart of an embodiment of the present invention.

FIG. 4 is another operation flowchart of the embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 magnetic disk control circuit 2 input / output control circuit 3 input / output circuit 4 magnetic disk 5 input / output inversion circuit 6 magnetic disk 7 comparison determination circuit 8 host device 31 write circuit 32 read circuit 51 write inversion circuit 52 read inversion circuit

Claims (1)

[Claims] 1. A magnetic disk controller for connecting a host device to first and second magnetic disks to control data input / output, and directly writes data sent from the host device to the first magnetic disk. Means, means for logically inverting and writing the data sent from the host device to the second magnetic disk, data written to the first magnetic disk and data written to the second magnetic disk. A magnetic disk control device comprising: a reading unit; and a comparison / determination unit that compares data of both systems and determines data that is logically inconsistent.