JPH04186422A - Duplex volume system - Google Patents

Abstract

PURPOSE:To improve processing performance by outputting a read request out of I/O instructions to a volume having less I/O instruction queues out of a master volume and a slave volume. CONSTITUTION:This duplex volume system is constituted of a duplex volume management table 7 having an I/O queuing matrix number counting area 8, an I/O signal queuing matrix number adding means 9, an I/O queuing matrix number subtracting means 10, a duplex volume 14 having the master and slave volumes, an I/O queue matrix 13, etc. A read request out of I/O instructions to the volume 14 is outputted to the master volume 11 or the slave volume 12 which has less I/O queues, i.e. the volume having lower load. Thus the processing performance can be improved.

Description

DETAILED DESCRIPTION OF THE INVENTION TECHNICAL FIELD The present invention relates to a duplex volume system, and more particularly to a volume duplex method for magnetic disks in which a magnetic disk is paired with two volumes, a primary volume and a secondary volume, to duplex a volume in a computer system.

Conventionally, in this type of magnetic disk volume duplication system, write commands among the input/output commands from the user program to the duplexed magnetic disks are issued to both the primary and secondary volumes, and the write commands are issued to both the primary and secondary volumes. It waits until all writes are completed, but a read command is issued correctly, and if a read failure occurs in the primary volume, it is issued to the secondary volume.

In this type of conventional volume duplication system, the secondary volume is the primary volume and is positioned as a standby system when a read failure occurs, even though the same data is stored in both the primary and secondary volumes. If the number of read accesses is large, the input/output commands will be concentrated on the positive volume and the load will not be reduced.

OBJECTS OF THE INVENTION The present invention has been made to eliminate the drawbacks of the conventional systems as described above, and an object of the present invention is to provide a dual volume system that can improve processing performance.

Configuration of the Invention The duplex volume system according to the present invention stores data of the same content in a first and second volume, and when an input/output command to the first and second volumes is a read command, A duplex volume system in which data is read from a volume, comprising first and second counting means for counting the number of input/output commands to each of the first and second volumes in a waiting state; 2
When the input/output command to the volume is a read command,
a comparison means for comparing the count values of the ml and the second counting means, and when the comparison means detects that the count value of the second counting means is smaller than the count value of the first counting means; , and control means for controlling reading of data from the second volume.

Embodiment Next, an embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention. In the figure, the duplex volume system according to one embodiment of the present invention includes a user program 1, input/output information management block creation means 2, input/output attribute determination means 3, input/output queue number comparison means 4, input/output Information management block 5
, a duplex volume management table 7 having an input/output device changing means 6 and a human output queue count area 8
, an input/output queue number adding means 9, an input/output queue number subtracting means 10, a duplex volume 14 having a primary volume 11 and a secondary volume 12, and an input/output queue 13.

2 is a flowchart showing the procedure of the input/output information management block creation means 2 and the input/output attribute determination means 3 of FIG. 1, and FIG. 3 is a flowchart showing the procedure of the input/output queue number comparison means 4 of FIG.
FIG.
The figure is a flowchart showing the operation of the human output queue number subtraction means 10 of FIG.

The input/output method of a duplex volume in which duplication is performed using the primary volume 11 and the secondary volume 12 will be described in detail using FIGS. 1 to 4.

First, when a read request to the primary volume 11 of the duplexed volume 14 of the user program 1 is issued, the input/output information block creation means 2 reads the input/output command and the device information of the primary volume 11, which is the input/output target device. The input/output information management block 5 containing the information is created (Step 2 in Figure 2).
1).

Next, the input/output attribute determination means 3 is activated, and refers to the contents of the input/output information management block 5 to determine whether the input/output target device is the duplex volume 14 or not. A table search is performed to determine whether the information is registered in the duplex volume management table 7 (step 22 in FIG. 2).

If it is an input/output target device or a non-duplex volume, normal input/output processing is started (steps 23 and 27 in FIG. 2).

If the input/output target device is the duplex volume 14 (step 23 in FIG. 2), it is determined whether it is a read request or a write request by referring to the input/output command of the input/output information management block 5 (step 23 in FIG. 2). 24).

As a result, if it is determined that the write request is for the duplex volume 14, duplex write processing is started (step 26 in FIG. 2).

When it is determined that the read request is for the duplex volume 14, the input/output queue number comparison means 4 is activated (step 25 in FIG. 2).

When the input/output attribute determining means 3 determines that the read request is to the duplex volume 14 and the input/output queue number ratio axis means 4 is activated, the input/output queue number comparing means 4 determines whether the read request is for the duplex volume 14 or the primary volume which is the input/output target device. 1] and the input/output queue count area 8 of each of the secondary volumes 12 are searched from the duplex volume management table 7 (step 31 in FIG. 3),
The values in the input/output queue count area 8 of each of the primary volume 11 and secondary volume 12 are compared.

As a result, if the value of the secondary volume 12 is small, the input/output device changing means 6 is started, and if the values of the primary volume 11 and the secondary volume 12 are the same, or if the value of the primary volume 11 is small, nothing is done, and Ota The power queue number adding means 9 is activated (step 32 in FIG. 3).

The input/output device changing unit 6 refers to the duplex volume management table 7 to obtain device information of the secondary volume 12 that is paired with the primary volume 11, and changes the target device information in the input/output information management block 5 to the primary volume 11. The information is changed from that to the information of the secondary volume 12 (step 33 in FIG. 3).

The values in the human output queue count area J area 8 compared by the input/output queue number comparing means 4 are added and subtracted by the input/output queue number adding means 9 and the input/output queue number subtracting means 10.

The human output queue number adding means 9 is the duplex volume management table 7 of the input/output target device of the input/output information management block 5.
Obtain the value of the input/output queue count area 8 (step 34 in Figure 3), and calculate the value of the count area 8 from the number of human output queues.
After adding 1 to the value of (step 35 in FIG. 3), the input/output information management block 5 is input to the input/output queue 13 (step 36 in FIG. 3).

When one of the input/output information management blocks 5 capable of issuing an input/output command is output from the input/output queue 13 for issuing an input/output command (Step 41 in FIG. 4), the Riki Ota queue number subtraction means 10 outputs an input/output command. Referring to the duplex volume management table 7, it determines whether the input/output target device information of the input/output information management block 5 that has been input/output is that of the duplex volume 14, and obtains the value of the input/output queue number count area 8 of the target device. (Step 42 in Figure 4).

If there is an input/output target device or duplex volume 14 (4th
(Step 43 in FIG. 4), 1 is subtracted from the value in the human output queue count area 8 of the target volume in the duplex volume management table 7 (Step 44 in FIG. 4).

After this, an input/output command issuing process for actually issuing an input/output command is started (step 45 in FIG. 4).

That is, the read request from the user program 1 is made by comparing the values in the input/output queue count area 8 of the primary volume 11 and the secondary volume 12 in the duplex volume management table 7. Issued to the volume with the smaller number of I/O queues.

In this way, read requests among input/output commands for the duplex volume 14 are issued to the one of the primary volume 11 and secondary volume 12 that is waiting for fewer input/output commands, that is, the one with a lighter load. It is possible to improve the performance of

DETAILED DESCRIPTION OF THE INVENTION According to the present invention, read requests among input/output commands for a duplexed volume are issued to the primary volume or the secondary volume, whichever is waiting for fewer input/output commands. This has the effect of improving performance.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, FIG. 2 is a flowchart showing the procedure of the human output information management block creation means and input/output attribute determination means shown in FIG. 1, and FIG.
The figure is a flowchart showing the operations of the input/output queue number comparing means, the input/output target device changing means, and the human output queue number addition means of FIG. 1, and FIG. 4 is the input/output queue number subtraction means of FIG. 1. 3 is a flowchart showing the operation of FIG. Explanation of symbols of main parts 2...Input/output information management block creation means 3...
...Input/output attribute determination means 4...Input/output queue number comparison means 6...
- Input/output target device changing means 8... Input/output queue number counting area 9... Input/output queue number adding means 10... Input/output queue number subtracting means 11・
...Primary volume 12 ...Secondary volume

Claims (1)

[Claims] (1) Duplication in which data with the same content is stored in the first and second volumes, and data is read from the first volume when an input/output command to the first and second volumes is a read command. A volume system,
first and second counting means for counting the number of input/output commands to each of the first and second volumes in a waiting state, and when the input/output command to the first and second volumes is a read command; , a comparison means for comparing the count values of each of the first and second counting means, and the comparison means detecting that the count value of the second counting means is smaller than the count value of the first counting means. and control means for controlling reading of data from the second volume when the volume is read from the second volume.