JPH0324691B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Technical Field of the Invention The present invention relates to replacement management in a disk device that alternately allocates a defective area on a disk to another area based on the contents of a replacement management table in a disk controller in the disk device. Regarding the method.

(2) Background of the technology In magnetic disk devices, for example, a defective area where information cannot be read or written due to a scratch on the disk surface is replaced with another good quality area for information to be read or written using alternate allocation. There is.
This alternate allocation is performed by a disk controller connected between the magnetic disk device and the central processing unit.

(3) Prior art and problems The conventional alternate area management method that controls the above-mentioned alternate allocation detects an abnormality in data by performing an access operation to a certain area on the disk, and then determines whether data is stored in that area. This is done by detecting a replacement address on the disk or accessing a predetermined location on the disk, that is, a replacement area, finding the target address, and reading/writing it.

However, the conventional method has the disadvantage that a total of two access operations are required when processing a defective region, that is, an abnormal region, and the processing time becomes extremely long.

In addition, a method has been proposed to solve the above problem by having a table in RAM on the controller that stores defective parts and new allocations for replacement processing (Japanese Patent Application Laid-Open No. 56-94454). After the allocation is made to the table in RAM, it is written to the table on the disk immediately before the device is stopped, so there was a problem that in the event of a power outage, the alternating processing that was in operation was invalidated and normal data could not be accessed.

There is also a problem in that even if the replacement process has been performed once and a replacement address has been assigned, if the area of the replacement record address becomes unusable due to scratches or the like, the replacement process cannot be performed again.

(4) Object of the Invention Therefore, in view of the problems of the conventional type described above, an object of the present invention is to provide a method for controlling the rotation management table on the disk or in the rotation management table storage device using the power source in the rotation management method in the disk device. Based on the concept of loading the data into the memory in the disk controller in advance at the time of input, etc., and performing the replacement process in both the memory and the table in the disk at the time of replacement processing, the processing time related to the abnormal part is shortened. To prevent problems even during a power outage and to enable alternate processing again.

(5) Structure of the Invention According to the present invention, it is an object of the present invention to detect defects on a disk in a replacement management method for a disk device in which defective parts on a disk are allocated and used on a replacement basis based on the contents of a replacement management table in a disk controller. The information in the replacement management table on the disk or replacement management table storage device to which the parts are allocated and used is loaded into the memory in the disk controller in advance when power is turned on to the disk device or during replacement processing, and The 0th and 1st bits indicating the presence or absence of replacement registration and the bits indicating information on abnormal parts are stored in each word in the area of one record of the replacement management table, and are stored on the disk each time a replacement block is allocated. Alternatively, a replacement registration operation is performed on the replacement management table on the replacement management table storage device and the replacement management table loaded into the memory, and the data is given according to the contents of the memory before accessing the disk device from the central processing unit, etc. The presence or absence of a replacement area is determined by checking the states of the 0th and 1st bits for the assigned address, and if alternate allocation has been made, the replacement address indicated in the memory contents is accessed, and if alternate allocation has been made. This is achieved by providing a replacement management method in a disk device, which is characterized in that the original address given is accessed if the original address is not available.

(6) Examples of the invention Examples of the invention will be described below with reference to the drawings. FIG. 1 shows a recording format on a disk in a disk device for implementing a method according to an embodiment of the present invention. This figure shows the format of recorded information for one track in one sector, and therefore one record. The recorded information is composed of a title part ID and a data part DT connected via a gap part G. The title ID is synchronized data
SYNC, an address mark AM indicating that the address data of the data section DT follows, address data consisting of a cylinder number CY, a head number HE, and a sector number SEC, and a check code CRC.
(Cycle redundancy check). The data portion DT includes synchronization data SYNC, a data mark DM indicating that data follows, data DATA, and a check code CRC.

When there is such an abnormal area on a disk containing recorded information, in the conventional replacement management system, the abnormal area is accessed once and the abnormality in the information is known, and then a process for performing replacement allocation is started. However, if the recording state of the abnormal region is unstable, it becomes impossible to accurately determine the presence or absence of an abnormality. To do this, for example, by making the ID section of the record initially determined to be an abnormal location unreadable, for example, the address mark AM is destroyed, it is possible to reliably detect that the record is an abnormal location. However, normally a disk controller only needs to read and write the data section DT and read the title section ID, and adding the function of writing the title section ID requires an increase in the amount of hardware.

Therefore, in the present invention, the information of the alternate area on the disk is loaded in advance into the general memory in the disk controller, and the presence or absence of an abnormality is detected before the disk is accessed. The need to add write functionality is removed.

FIG. 2 shows the configuration of an information processing system including a disk device for implementing the method of the present invention.
In the figure, 1 is a central processing unit, 2 is a main storage device, 3 is a bus line, 4 is a disk controller, and 5 is a main storage device.
is a disk device. disk controller 4
is connected between the bus line 3 to which the central processing unit 1, the main storage device 2, etc. are connected, and the disk device 5, and performs processing related to replacement management according to the present invention, retry of disk processing, and control of other operations. It is something. Such a disk controller 4 includes a bus line controller 6, an input/output processor 7, a memory 8, and a disk interface 9.

Further, FIG. 3 shows a replacement management table provided on the disk of the disk device 5 described above. The table in the figure shows, for example, that the disk device described above has 100 cylinders, 4 heads and therefore 4 recording surfaces, 1 track has 16 sectors, or 16 records, and 1 record. is 256
This assumes a case consisting of bytes. Third
For example, the table shown in the figure is located in cylinder 0,
One is provided for each head. Each table has 1
1st to 1st in the record (256 bytes) area
It is composed of 15 words, and the 16th to 127th words are unused areas. Note that one word consists of 16 bits, that is, 2 bytes. cylinder 0
Each of records 1 to 15 serves as a replacement area, and the record numbers 1 to 15 are associated with the first to 15th words of the replacement management table. Note that the 0th word of the replacement management table contains a hexadecimal value to indicate that record 0, that is, sector 0, is the replacement management table and cannot be used.
8000 is written.

Each word in the replacement management table is structured as shown in the diagram on the right side of Figure 3, and the 0th and 1st bits indicate whether or not a replacement has been registered.When these bits are 00, it means that a replacement has not been registered. 01 (the first bit is 1) indicates that the area is unused, and 10 and 11 indicate that it is used for other purposes. Also,
The 2nd to 7th bits and the 8th to 15th bits contain the record (sector) number SEC of the abnormal area, respectively.
and cylinder number are stored.

A replacement management method in a disk device having a replacement management table as described above will be explained with reference to FIGS. 2 and 3. In the system shown in FIG. 2, when the power of the system is turned on or when an abnormal part is discovered during system operation and a replacement is registered in the replacement management table, the disk controller 4 performs replacement management on the disk. The contents of the table are read and loaded into the memory 8. In this state, when a read or write command is sent from the central processing unit 1 to the disk controller 4 via the bus line 3, the input/output processor 7 uses the alternate management information loaded in the memory 8 before accessing the disk device 5. Search for a table. In other words, a read or write command contains the address on the disk (head number, cylinder number, record number) to be read or written, but the rotation management corresponding to the head number in this given address is The contents of the 0th and 1st bits of each word in the table are determined to check whether replacement registration has been performed. If replacement registration has been performed, i.e. 0th and 1st
If there is a word whose bit contents are 00, the record number and sector number in the address are compared with the contents of the second to fifteenth bits of the word. If the two match, a record with a number corresponding to the word number is prepared as a replacement area, and the disk controller replaces the address given by the command from the central processing unit 1. to access the replacement area. In this case, access to the abnormal site is not necessary.

Furthermore, if the contents of the 0th bit and the 1st bit are 01, it indicates that there is no replacement processing, and if the contents of the 0th bit and the 1st bit are 10, it indicates that the corresponding replacement area cannot be used. . In this way, in the present invention, the replacement management table has a flag area in the 0th bit and the 1st bit, so that re-registration is possible.

(7) Effects of the Invention As explained above, according to the present invention, it is not necessary to access the abnormal part and the replacement management table on the disk when executing a read or write command, so the processing time regarding the abnormal part is reduced. Since there is no need to rewrite the contents of the ID section of the abnormal area, it is possible to prevent the hardware from becoming complicated, and even in the event of a power outage, it is possible to perform alternate processing again without any problems.

[Brief explanation of the drawing]

FIG. 1 is a format diagram showing an information device on a disk in a disk device for implementing a method according to an embodiment of the present invention, and FIG. 2 is a format diagram showing an information device on a disk for implementing a method according to an embodiment of the present invention. FIG. 3 is a block circuit diagram showing an information processing system including a disk device, and FIG. 3 is an explanatory diagram showing the configuration of a replacement management table. 1...Central processing unit, 2...Main storage device, 3...
...Bus line, 4...Disk controller, 5...
Disk device, 6... Bus line controller, 7...
...I/O processor, 8...Memory, 9...Disk interface.

Claims (1)

[Claims] 1. In a disk device replacement management method in which defective parts on a disk are allocated and used based on the contents of a replacement management table in a disk controller, information in the replacement management table on the disk or in the replacement management table storage device is used. The 0th and 0th rows are loaded in advance into the memory of the disk controller when the power is turned on for a disk device or the like or during replacement processing, and the presence or absence of replacement registration is displayed in each word in an area in one record of the replacement management table. 1 bit and a bit indicating information on the abnormal part are stored, and each time a replacement block is allocated, replacement is registered in the replacement management table on the disk or replacement management table storage device and the replacement management table loaded in the memory. Before accessing the disk device from the central processing unit, etc., the presence or absence of a replacement area is determined by checking the states of the 0th and 1st bits for the address given by the contents of the memory, and the replacement area is determined. 1. A replacement management method in a disk device, characterized in that if the replacement address is allocated, the replacement address indicated in the memory contents is accessed, and if the replacement address is not allocated, the given original address is accessed.