JPH04311217A - External storage controller - Google Patents

Abstract

PURPOSE:To speed up updated data writing operations when each data recorded at the logically same position are updated by means of an external storage controller which controls plural external storage devices storing check codes for correcting data and errors. CONSTITUTION:A CPU 6 stores updated data stored in a RAM 7 at the logical position to be updated of disk drives DD-1 to DD-n for data. The CPU 6 stores a new check code at the logical positions to be updated of disk drives CD-1 to CD-m for checking which are delayed by prescribed accessing time in corresponding to the processing time from the generation of a new check code to the writing of the new code against the logical position at which the data of the disk drives DD-1 to DD-n are recorded.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an external storage control device used, for example, in an external storage system having a plurality of hard disk drives.

0002

2. Description of the Related Art Conventionally, there are external storage systems that use a plurality of external storage devices, such as disk array devices. This system consists of a host computer, multiple hard disk drives (HD
D) and an external storage control device (controller).

[0003] A disk array device is an external storage device having a logically identical configuration so that data can be simultaneously read/written to the same logical location on a plurality of HDDs. Specifically, the controller can simultaneously read data in the same sector (same logical sector) from each HDD, and write data in the same sector to update the data.

By the way, normally, when an error occurs in each data recorded on each HDD (data HDD), a check HDD (usually a plurality of HDDs) is provided to store a check code for error correction. It is being The controller generates a check code based on each piece of data stored in the same logical location on the data HDD, and stores it in the check HDD that is in the same logical location as the data HDD.

[0005] In such an external storage system, data H is
When updating recorded data on a DD, the controller reads old data, which is recorded data, from the logical sector to be updated. In addition, the controller is a check HDD.
The check code of the same logical sector as the logical sector to be updated is read from. The controller generates a new check code based on the updated data. Then, the updated data is stored in the logical sector to be updated on the data HDD, and a new check code is stored in the same logical sector on the check HDD.

[0006]

[Problems to be Solved by the Invention] In the conventional method, when updating recorded data on a data HDD, the old data and check code to be updated must be read before the new updated data and check code are stored. The process is being executed. For this reason, processing time is required to read out the old data and check code to be updated, which takes a long processing time to complete the update, which is a factor in the slow data write operation.

An object of the present invention is to update each piece of data recorded at the same logical location in an external storage control device that controls a plurality of external storage devices that store data and check codes for error correction. As a result, the purpose is to speed up the write operation of update data.

[0008]

[Means for Solving the Problems] The present invention simultaneously reads/writes each piece of data at the same logical location to a plurality of external storage devices for data, and simultaneously reads and writes data at the same location logically to each external storage device for data. Memory means for temporarily storing updated data and new check codes in an external storage control device that controls reading/writing of check codes for error correction corresponding to each data in an external storage device for checking a configuration; A check code generating means for generating a new check code based on the updated data, storing the new check code in an external storage device for checking, and placing the updated data in a logical position to be updated in each external storage device for data. This device has data write control means for storing data.

[0009]

According to the present invention, the data write control means stores the update data stored in the memory means in the logical location to be updated in each external data storage device. and generates a new check code generated by the check code generation means for the logical position where the data is recorded in the external storage device for data, depending on the processing time until it is written. The data is stored at a logical location in the external storage device for checking delayed by a predetermined access time.

0010

DESCRIPTION OF THE PREFERRED EMBODIMENTS Examples of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an external storage system according to the same embodiment. This system consists of a host computer 1, an external storage controller 2, a plurality of data disk drives DD-1 to DD-n, and a plurality of check disk drives CD-1 to CD-m, which constitute the main body of the computer. has.

[0012] Data disk drives DD-1 to DD
-n is an external storage device each having a recording medium consisting of a hard disk and storing data such as user data other than the check code. Check disk drives CD-1 to CD-m are data disk drives D.
It is an external storage device that has the same logical configuration as D-1 to DD-n and stores error correction check codes corresponding to each data.

The controller 2 has a host interface 3 and is connected to the host computer 1 to exchange data, commands, and the like. On the other hand, controller 2 is
Data disk interfaces 4-1 to 4-n and check disk interfaces 5-1 to 5-m
has. Interfaces 4-1 to 4-n are connected to data disk drives DD-1 to DD-n, respectively, and exchange data according to read/write operations. Further, the interfaces 5-1 to 5-m are connected to the check disk drives CD-1 to CD-m, respectively, and exchange check code data according to read/write operations.

The controller 2 also includes a CPU 6 that controls the entire controller, a RAM 7 that stores various data necessary for data processing by the CPU 6, and a ROM that stores a program that determines the content of data processing by the CPU 6.
It has 8.

The CPU 6 is a control means including check code generation means and data write control means of the present invention, and is composed of a microprocessor. A read/write RAM 7 stores update data and read data for the data disk drives DD-1 to DD-n, and stores new check codes and read check codes for the check disk drives CD-1 to CD-m. It's memory. In addition, instead of RAM7, CPU
A plurality of registers provided inside 6 may be used. Next, the operation of this embodiment will be explained.

First, as a basic operation, the controller 2 controls the data disk drives DD-1 to D in response to read/write commands from the host computer 1.
The same logical position (logical sector) of D-n is accessed simultaneously. That is, the CPU 6 uses the host interface 3
When a write command is received from, for example, the host computer 1 through the host computer 1, the write data transferred from the host computer 1 is temporarily stored in the RAM 7. C.P.
U6 accesses the same logical sector of the data disk drives DD-1 to DD-n through the interfaces 4-1 to 4-n, and writes each data stored in the RAM 7. That is, as shown in FIG. 4, each data DL is written to, for example, each logical sector L of the data disk drives DD-1 to DD-n.

On the other hand, the CPU 6 generates a check code CL based on the write data DL, and stores it in the check disk drives CD-1 to CD-m through the interfaces 5-1 to 5-m. At this time, as shown in FIG. 4, the check code CL is applied to a logical sector (L+K
) is stored in the location.

Here, each disk drive (DD-1 to
DD-n, CD-1 to CD-m) disk rotation speed is R.
, when the number of logical sectors in one track of each disk is S, the delay time T for the number of K sectors is "T = (K-1)
/(R×S)”.

That is, data disk drive DD-1
~ After writing (or reading) data DL to logical sector L of DD-n, the check code CL corresponding to that data is sent to the disk drive CD for checking.
A delay time T occurs before writing to the logical sector (L+K) of -1 to CD-m. This delay time T
is set to be a time equivalent to or longer than the processing time required for the CPU 6 to generate a new check code and write it to the check disk drives CD-1 to CD-m.

[0020] Also, check disk drive CD-
The check code (for example, CL) of each same logical sector stored in CD-1 to CD-m is used to simultaneously detect errors in a predetermined number of E drives or less among the data disk drives DD-1 to DD-n. It is data that has the ability to be corrected. Under these basic conditions, the data update operation for the data disk drives DD-1 to DD-n will be explained.

As shown in step S1 in FIG. 2, upon receiving a data update command from the host computer 1, the CPU 6 updates the data disk drives DD-1 to DD.
- Find the disk drive to be updated in n (
Step S2). In this case, the following processing differs depending on whether the number of data disk drives to be updated is equal to or less than E-1 that can simultaneously correct errors.

If the number is E-1 or less (YES in step S3), the CPU 6 temporarily stores the update data transferred from the host computer 1 in the RAM 7 (step S4). The CPU 6 writes the update data stored in the RAM 7 to the same logical sector of the data disk drives DD-1 to DD-n to be updated, and updates the old recorded data with the new write data (step S5).

Next, the CPU 6 accesses the non-updated logical sector which is the same sector as the updated logical sector, reads out each piece of recorded data, and stores it in the RAM 7 (step S6). The CPU 6 generates a new check code corresponding to all recorded data in the same logical sector based on the update data stored in the RAM 7 and the recorded data that is not to be updated, and sends the check code to the check disk drives CD-1 to CD-m.
(steps S12, S13).

At this time, as shown in FIG. 4, the check code CL indicates that the data disk drives DD-1 to D
It is stored in the position of the logical sector (L+K) which is temporally delayed by the number of K sectors with respect to the logical sector L of D-n. Therefore, from the time the CPU 6 accesses the data disk drives DD-1 to DD-n that are not subject to update,
As described above, there is a delay time T before accessing the same logical sector in the check disk drives CD-1 to CD-m. Thereby, the CPU 6 can generate a new check code using the delay time T and store it in the corresponding logical sector of the check disk drives CD-1 to CD-m. In other words, new check codes can be continuously generated and stored after accessing the non-updated data disk drives DD-1 to DD-n, resulting in a significant increase in check code generation processing time. It can be shortened to

On the other hand, after the update data is once stored in the RAM 7, the data disk drive DD-1 to be updated is
~ DD-n. Therefore, when the CPU 6 generates a new check code, it is not necessary to access all the logical sectors of the data disk drives DD-1 to DD-n that are to be updated and not to be updated; All you need to do is access.

Next, if the number of data disk drives to be updated is E or more (NO in step S3).
, the CPU 6 stores the updated data in the RAM 7, and then
All logical sectors to be updated are accessed, old data is read out and stored in the RAM 7 (steps S7, S8). The CPU 6 performs an error check on the old data of all logical sectors stored in the RAM 7 (step S9).

[0027] As a result of this error check, if there is an error in the old data (YES in step S10), the CP
U6 is a check disk drive CD-1 to CD-
Read the check code from the corresponding logical sector of m,
Execute the process to correct old data with errors (
Step S11).

If there is no error in the old data (NO in step S10), the CPU 6 creates a new check code corresponding to all recorded data in the same logical sector based on the updated data stored in the RAM 7 and the old data. generate,
Store in the corresponding logical sector of the check disk drives CD-1 to CD-m (steps S12, S13)
. Thereafter, the CPU 6 waits for a time corresponding to the delay time T, and then updates the data disk drive D to be updated.
A process of accessing the logical sectors D-1 to DD-n to be updated and writing the update data stored in the RAM 7 is executed.

In this way, in the case of E or more disk drives for which it is impossible to perform error correction at once, all logical sectors to be updated are checked using the check codes of the check disk drives CD-1 to CD-m. It accesses the old data and executes the process of reading it and storing it in the RAM 7. After this, if old data in which an error has occurred exists, a check code is read from the corresponding logical sector and error correction processing is executed.

[0030]

As described in detail above, according to the present invention, in an external storage control device that controls a plurality of external storage devices that store data and check codes for error correction,
When updating data recorded in the same logical location, the updated data can be written to the same location in each device to be updated without reading old data to be updated. In this case, new check codes can be continuously generated and stored based on update data and non-update target data. Therefore, it is possible to speed up the write operation of update data as a result.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing main parts of an external storage system according to an embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of the embodiment.

FIG. 3 is a flowchart for explaining the operation of the embodiment.

FIG. 4 is a conceptual diagram for explaining the operation of the embodiment.

[Explanation of symbols]

1...Host computer, 2...Controller, 6...CP
U, 7...RAM, DD-1~DD-n...Data disk drive, CD-1~CD-m...Check disk drive.

Claims (1)

[Claims] 1. An external storage device for checking that simultaneously reads/writes each piece of data at the same logical location to a plurality of external storage devices for data, and that has the same configuration logically as each of the external storage devices for data. In an external storage control device that controls reading/writing of check codes for error correction corresponding to each of the above-mentioned data, the update data and new check codes for updating part or all of the above-mentioned data are temporarily stored. a new check code using only the updated data stored in the memory means or both the updated data and non-updated data read from the external storage device for each data; a check code generating means for generating a check code and storing it in the memory means; and a check code generating means for storing the update data stored in the memory means in a logical position to be updated in each of the external storage devices for data, and by the check code generating means. A predetermined access time corresponding to the processing time to generate and write the new check code to the logical position where the data of the data external storage device is recorded. an external storage control device characterized by comprising: data write control means for storing data in a logical position of the external storage device for checking that is delayed by a certain amount of time.