JPH04215142A - Duplex disk control system - Google Patents

Abstract

PURPOSE:To frequently copy stored information of a current disk device to a standby disk device without degrading the processing capability and to extract stored information of the current disk device even in the case of abnormality of a current disk controller in the duplex disk control system for a duplex information processing system having a standby auxiliary system. CONSTITUTION:A shared disk device 8 where data is written and read by both systems is provided, and current and standby disk controllers 5 can copy stored information of a current disk device 6 to a standby disk device 6 through the shared disk device 8; and if the current disk controller 5 is abnormal, the current disk device 6 is connected to the standby disk controller 5 by switching to extract stored information of the current disk device 6 through the standby disk controller 5.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention consists of a dual system each equipped with a central processing unit, a channel control unit, a disk control unit, and a disk device, one system operates as an active system, and the other system operates as an active system. This invention relates to a duplex disk control method in an information processing system in which a disk is on standby as a standby system.

0002

2. Description of the Related Art FIG. 6 is a diagram showing an example of a conventional information processing system. In FIG. 6, the information processing system is
Central control unit (CC) 1, main memory unit (MM) respectively
2, a dual system comprising a channel controller (CHC) 3, a disk controller (DKC) 5, and a disk device (DK) 6 (each system is referred to as the 0 system and 1 system;
-0 and -1 are added to each device constituting the system), one system (e.g. 0 system) operates as the active system, and the other system (e.g. 1 system) operates as the backup system. stand by.

[0003] When the active system stops operating, the standby system immediately starts operating as the active system, and in order to shorten the period during which the entire information processing system is out of operation as much as possible, it is necessary to install a standby disk unit (DK) 6-1. The current disk unit (DK) 6-0
It is necessary to hold the same memory information as .

[0004] Therefore, after initially starting operation of only the active system (0 system), the active central controller (CC) 1-0 extracts a predetermined amount of data from the active disk drive (DK) 6-0 between operations. Storage information is stored in a channel controller (CHC) 3-0, a channel bus 4-0, and a disk controller (DKC) 5-.
0, and once stored in the main memory (MM) 2-0, the channel controller (CHC) of the standby system (1 system)
) 3-1, a spare disk unit (DK) 6 via a channel bus 4-1 and a disk controller (DKC) 5-1.
-1 is stored in the corresponding storage area.

[0005] The above process is carried out using a disk device (DK) 6-
By executing this on all the storage information of 0 and 6-1, the same storage information as the active disk device (DK) 6-0 will be copied to the spare disk device (DK) 6-1. .

The current central controller (CC) 1-0 stores information stored in the disk devices (DK) 6-0 at the time when the information processing system starts operating.
) After copying to DK 6-1, repeat the same copy periodically to keep the stored information in both disk devices (DK) 6-0 and 6-1 as consistent as possible over the entire operation period. ing.

However, with the expansion of the scale of information processing systems and the increase in the capacity of disk devices (DKs) 6, the amount of information to be copied has also increased, and the time required for copying has also continued to increase. Repeating copying becomes impossible because it puts pressure on the processing capacity of the active central controller (CC) 1-0, and as a result, as the copying cycle becomes longer, the information stored in the spare disk unit (DK) 6-1 and Current disk device (D
K) The discrepancy with the stored information of 6-0 becomes significant.

[0008] In such an information processing system, when a failure occurs in the active disk controller (DKC) 5-0,
Even if the storage information of the active disk unit (DK) 6-0 cannot be extracted and the active system is switched from the 0 system to the 1 system, the storage information of the backup disk unit (DK) 6-1 is not necessarily extracted from the active disk unit (DK) 6-1. ) 6-0, and the operation of the information processing system will be interrupted.

[0009]

[Problems to be Solved by the Invention] As is clear from the above description, in a conventional information processing system, information stored in the active disk device (DK) 6-0 is copied to the spare disk device (DK) 6-1. In order to
Since the storage was sequentially stored in the spare disk device (DK) 6-1 via the disk device (DK) 6-1, a large amount of the processing capacity of the active central controller (CC) 1-0 was used.
With the increase in the amount of stored information, the copying cycle has to be extended, and there is a problem in that it is difficult to match the stored information in the disk devices (DKs) 6 of both systems.

In such a state, the current disk control device (D
If KC) 5-0 is affected, the current disk unit (DK)
6-0 information cannot be extracted, and the spare disk device (
There was a problem in that the information processing system had to be restarted using the information stored in the DK) 6-1 that was not guaranteed to be completely consistent, which impaired the reliability of the information processing system.

In order to prevent such a decrease in reliability, the disk units (DKs) 6-0 and 6-1 of both systems are connected by a confounding line 7.
directly connected to the active disk controller (DKC)5.
-0 transfers all information stored in the active disk unit (DK) 6-0 to the spare disk unit (DK) via the confounding line 7.
) 6-1 in parallel, but a failure that occurs in one disk device (DK) 6 will have an adverse effect on the other disk device (DK) 6, causing the information processing system to This will lead to a decrease in reliability and is not an effective solution.

[0012] The present invention provides an information processing system consisting of a dual system each comprising a central processing unit, a channel control device, a disk control device, and a disk device, in which information stored in an active disk device is stored in the information processing system. The purpose is to enable copying to backup disk devices as frequently as possible without putting pressure on the processing capacity of the processing system.
Another object of the present invention is to make it possible to extract the storage information of the current disk device even if the current disk control device is affected.

[0013]

[Means for Solving the Problems] Fig. 1 is a diagram showing the principle of the present invention. Fig. 1 (a) shows the principle of the present invention (Claim 1), and Fig. 1 (b) shows the principle of the present invention (Claim 1). The principle of 2) is shown.

In FIG. 1, 10 is a central processing unit, 3 is a channel control device, 5 is a disk control device, and 6 is a disk device. Two sets of each device are provided, forming a dual system, with one operating as an active system and the other on standby as a standby system.

Reference numeral 8 denotes a shared disk device provided according to the present invention (claim 1). 100 is a first connection means provided in the disk control device 5 according to the present invention (claims 1 and 2).

Reference numeral 200 denotes second connection means provided in the disk control device 5 according to the present invention (claim 1). 3
00 is a disk control device 5 according to the present invention (claim 1).
This is a copying means provided in the.

Reference numeral 400 denotes third connection means provided in the disk control device 5 according to the present invention (claim 2). 5
00 is the disk control device 5 according to the present invention (claim 2).
This is a switching means provided in the.

[0018]

[Operation] The shared disk device 8 allows writing and reading from each system. The first connection means 100 controls the disk devices 6 in the same system independently of the disk devices 6 and shared disk devices 8 in other systems.

The second connection means 200 controls the shared disk device 8 independently of each disk device 6. The copying means 300 copies information stored in one of the disk devices 6 controlled by the first connecting means 100 and the shared disk device 8 controlled by the second connecting means 200 to the other. do.

The third connection means 400 controls the disk devices 6 in other systems independently of the disk devices 6 in the same system. The switching means 500 connects the disk device 6 in the same system to the first connection means 100 of the disk control device 5 in the same system while the disk control device 5 in the same system is operating normally;
When the disk control device 5 in the same system does not operate normally, the disk device 6 in the same system is switched and connected to the third connection means 400 of the disk control device 5 in the other system.

[0021] The normality of each disk control device 5 can be monitored within each disk control device 5, and the monitoring result can be notified to the switching means 500 within the same system, or the normality can be monitored by the switching means 500 within the same system. is taken into account.

It is also considered that the power supply for each disk control device 5 is provided independently from the power supply for the disk device 6 and the switching means 500 for each system. Therefore, according to the present invention (claim 1), the active disk control device copies the storage information of the active disk device to the shared disk device, and the spare disk control device copies the storage information of the shared disk device to the spare disk device. Therefore, the storage information of the current disk device can be copied to the spare disk device without putting pressure on the processing capacity of the current central processing unit. Even if the device does not operate normally, the information stored in the active disk device can be extracted via the spare disk control device, and the reliability of the information processing system is greatly improved.

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a diagram showing an information processing system according to an embodiment of the present invention (claim 1), FIG. 3 is a diagram showing an example of the disk control device in FIG. 2, and FIG. 4 is a diagram showing an example of the disk control device in FIG. ) is a diagram showing an information processing system according to an embodiment of the present invention, and FIG. 5 is a diagram showing an example of the bus switching device in FIG. 4. Note that the same reference numerals indicate the same objects throughout the figures. Also in FIGS. 2 and 4, the information processing system includes a central controller (CC) 1 and a main memory (MM), respectively.
) 2, a channel controller (CHC) 3, a disk controller (DKC) 5, and a disk device (DK) 6. One of the systems (for example, the 0 system) serves as the active system. The other system (for example, the 1st system) is on standby as a standby system.

In FIG. 2, a central control unit (CC) 1 and a main memory (MM) 2 are shown as the central processing unit 10 in FIG. 1(a), and a first connection means in FIG. 1(a) is shown. 100 and a second connection means 200, disk connection control units (DIF) 55 and 56 are provided in each disk controller (DKC) 5, and as shown in FIG.
In this case, the copy unit (CPY) 521 as the copying means 300 in FIG. 1(a) is a read-only memory (ROM).
52 and a buffer memory (BFM)
57 are provided.

Note that the disk connection control units (DIF) 55 and 56 in each disk controller (DKC) 5 independently control the disk device (DK) 6 and shared disk device (CDK) 8 in the same system. Therefore, there is no risk that a failure in one will spread to the other.

In FIGS. 2 and 3, the initially active system (0
After starting operation of only the current central control unit (CC system)
) 1-0 to the current disk controller (DKC) 5-0,
The specified amount of storage information stored in the specified storage area of the disk device (DK) 6-0 is shared with the disk device (CDK).
The current channel control device (CH
C) 3-0 and the working channel bus 4-0.

In the active disk controller (DKC) 5-0 that has received the copy command, the processing unit (MPU) 51
-0 is stored in the read-only memory (ROM) 52-0 by executing the copy unit (CPY) 521-0, the buffer memory (BFM) 57-0 and the disk connection control unit (DIF) 55-0. After extracting a specified amount of storage information from the specified storage area of the active disk unit (DK) 6-0 via the DK 0 and storing it once in the buffer memory (BFM) 57-0, the disk connection control unit (DIF) 56-0 0 to the corresponding storage area of the shared disk device (CDK) 8.

The processing unit (MPU) 51-0 stores the storage information specified by the copy command in the current disk unit (DK).
When the transfer from 6-0 to the shared disk device (CDK) 8 is completed, the channel bus 4-0 and the channel control device (CDK)
Current central control unit (CC) 1-0 via HC) 3-0
Then, a copy completion notification in response to the copy command is sent back.

[0029] The active central controller (CC) 1-0 that has received the copy completion notification sends a backup disk controller (DKC)
5-1, the backup channel controller (CHC) sends a copy command to transfer a specified amount of storage information stored in the designated storage area of the shared disk device (CDK) 8 to the backup disk device (DK) 6-1. 3-1 and spare channel bus 4-
1.

In the spare disk controller (DKC) 5-1 that received the copy command, the processing unit (MPU) 51
-1 is stored in the read-only memory (ROM) 52-1 by executing the copy unit (CPY) 521-1, the buffer memory (BFM) 57-1 and the disk connection control unit (DIF) 56-1 are stored in the read-only memory (ROM) 52-1. After extracting a specified amount of storage information from the specified storage area of the shared disk device (CDK) 8 via the shared disk device (CDK) 1 and storing it once in the buffer memory (BFM) 57-1, the disk connection control unit (DIF) 55-1 is extracted. The data is stored in the corresponding storage area of the spare disk device (DK) 6-1 via the data storage system.

The processing unit (MPU) 51-1 transfers the storage information specified by the copy command to a shared disk device (CDK).
) 8 to the spare disk unit (DK) 6-1, the channel bus 4-1 and channel control device (C
Current central control unit (CC) 1-0 via HC) 3-1
Then, a copy completion notification in response to the copy command is sent back.

The active central controller (CC) 1-0 repeatedly transmits the above-mentioned copy command to the active disk controller (DKC) 5-0 and the backup disk controller (DKC) 5-1. Control device (DKC) 5
-0 and the spare disk controller (DKC) 5-1,
All storage information of the active disk device (DK) 6-0 is transferred to the spare disk device (DK) 8 via the shared disk device (CDK) 8.
Transfer completed to DK) 6-1.

The current central controller (CC) 1-0 stores information stored in the disk devices (DK) 6-0 at the time when the information processing system starts operating.
) After copying to DK 6-1, repeat the same copy periodically to keep the stored information in both disk devices (DK) 6-0 and 6-1 as consistent as possible over the entire operation period. ing.

As is clear from the above description, according to this embodiment, the active central controller (CC) 1-0 includes the active disk controller (DKC) 5-0 and the spare disk controller (DKC) 5-1. By simply transmitting the above-mentioned copy command to the active disk controller (DKC) 5-0 and the backup disk controller (DKC) 5-1, the entire storage information of the active disk controller (DK) 6-0 is shared. Disk device (
Since the data is transferred to the spare disk device (DK) 6-1 via the CDK) 8, the storage information of the spare disk device (DK) 6-1 should match the storage information of the active disk device (DK) 6-0 as much as possible. Even if the copying cycle is shortened, there is no possibility that the processing capacity of the current central control unit (CC) 1-0 will be overwhelmed.

Note that the shared disk device (CDK) 8 can also be used for the purpose of expanding the storage capacity of each disk device (DK) 6; for example, when the storage capacity of the disk device (DK) 6-0 is temporarily insufficient. In this case, the excess information is temporarily stored in the shared disk device (CDK) 8, and later when the storage capacity of the disk device (DK) 6-0 becomes available, the information is stored in the shared disk device (CDK) 8. It is also possible to copy the excess information stored in 8 to the disk device (DK) 6-0.

Next, in FIGS. 4 and 5, FIG.
) as the central processing unit 10 in the central control unit (C
C) 1 and main memory (MM) 2 are shown, and also in FIG.
(b) Disk connection control section (DIF) 55 as the first connection means 100 and the third connection means 400
and 58 are provided in each disk controller (DKC) 5, and a bus switching device (BSW) 9 is provided as the switching means 500 in FIG. 1(b).

Note that the disk connection control units (DIF) 55 and 58 in each disk controller (DKC) 5 independently connect the disk units (DK) 6 in the same system and the disk units (DK) 6 in other systems, respectively. Since the systems are controlled in the same way, there is no risk that a failure in one will spread to the other.

Further, in FIG. 5, each disk controller (DKC) 5 is provided with a failure detection section 59, and each bus switching device (BSW) 9 is provided with a relay control section 91 and a relay (RY). ) 92 are provided.

Each failure detection unit 59 determines whether or not its own disk controller (DKC) 5 is operating normally.
The operating status of each component circuit or power supply status is monitored, and if it is determined that it is operating normally, a failure is transmitted to the bus switching device (BSW) 9 in the same system. The detection signal dt is set to a normal state (for example, logic “
0"), and if it is determined that the bus is not operating normally, the fault detection signal dt transmitted to the corresponding bus switching device (BSW) 9 is set to an abnormal state (for example, logic "1").
).

Each relay control section 91 receives a fault detection signal dt transmitted from the fault detection section 59 in the same system.
is monitored, and the failure detection signal dt is in a normal state (logic “0”)
is set, the same bus switching device (BSW)
) 9 is maintained in an operating state, but if the failure detection signal dt is set to an abnormal state (logic "1"), the relay (RY) 92 is restored.

While each relay (RY) 92 is operating, each disk device (DK) 6 is connected to the operating relay contact (RY).
) 93 (dotted line state in FIG. 5) to the disk connection control unit (DIF) 55 of the disk controller (DKC) 5 in the same system, but the relay (R
When the Y) 92 is restored, each disk device (DK) 6 connects the disk of the disk controller (DKC) 5 in the other system via the relay contact (ry) 93 (solid line state in FIG. 5) in the restored state. It is switched and connected to a connection control unit (DIF) 58.

In FIGS. 4 and 5, a working disk controller (DKC) 5-0 and a spare disk controller (DKC) 5-0 are shown.
DKC) 5-1 are operating normally, the fault detection units 59-0 and 59-1 in the active disk controller (DKC) 5-0 and standby disk controller (DKC) 5-1 The output fault detection signal dt is set to the normal state (logic "0"), and the relay (R) in each bus switching device (BSW) 9-0 and 9-1
Y) 92-0 and 92-1 maintain the operating state, and the active disk drive (DK) 6-0 has a relay contact (ry).
) 93-0 to the active disk controller (DKC) 5
-0 is connected to the disk connection control unit (DIF) 55-0, and the spare disk device (DK) 6-1 is connected to the spare disk control device (DK) 55-0 via the relay contact (ry) 93-1.
C) is connected to the disk connection control unit (DIF) 55-1 of 5-1, and the active central controller (CC) 1-0 is
Active channel controller (CHC) 3-0, active channel bus 4-0, and active disk controller (DKC) 5-
0, the storage information of the active disk unit (DK) 6-0 is written and read, and the spare channel controller (CHC) 3-1, the spare channel bus 4-1, and the spare disk control device are used as necessary. Information stored in the spare disk device (DK) 6-1 can also be written and read via the device (DKC) 5-1.

In this state, the current disk control device (D
For example, if a failure occurs in the KC) 5-0 and it no longer operates normally, the failure detection unit 59-0 detects the own disk controller (D
The relay control unit in the bus switching device (BSW) 9-0 determines that the KC) 5-0 is not operating normally and sets the output failure detection signal dt to an abnormal state (logic "1"). 91-0 restores relay (RY) 92-0.

As a result, the current disk device (DK) 6-
0 is connected to the disk connection control unit (DIF) 58-1 of the spare disk controller (DKC) 5-1 via the restored relay contact (ry) 93-0.

It should be noted that the spare disk controller (DKC) 5-
1 is operating normally, so the spare disk unit (DK
) 6-1 is the relay contact (ry
) 93-1 to the spare disk controller (DKC) 5
However, as mentioned above, the spare disk controller (DKC) 5-1 is connected to the active disk device (DK) 6-0 and the spare disk device (DK) 55-1. ) 6-1 respectively independently,
There is no risk that a failure in one will spread to the other.

In such a state, the current central control unit (CC
) 1-0 stores the storage information of the active disk device (DK) 6-0, the spare channel controller (CHC) 3-1, the spare channel bus 4-1, and the spare disk controller (DKC) 5-.
Writing and reading are possible via the disk connection control unit (DIF) 58-1 of No. 1 and the relay contact (ry) 93-0 in the recovery state of the active bus switching device (BSW) 9-0.

As is clear from the above description, according to this embodiment, even if the active disk controller (DKC) 5-0 does not operate normally, the active central controller (CC) 1-
0 can immediately write and read the storage information of the active disk unit (DK) 6-0 via the spare disk controller (DKC) 5-1, so the information processing system can operate in a short time. It is possible to restart.

It should be noted that FIGS. 2 to 5 are only one embodiment of the present invention (claims 1 and 2), and for example, the normality of each disk controller (DKC) 5 is determined by the built-in failure detection unit. It is not limited to what is monitored by 59,
Many other modifications may be considered, such as monitoring by a bus switching device (BSW) 9 within the same system, but the effects of the present invention remain the same in either case. Furthermore, the information processing systems shown in FIGS. 2 and 4 are not limited to being constructed separately, and many other modifications may be considered, such as combining the functions of both information processing systems. In either case, the effects of the present invention remain the same.

[0049]

As described above, according to the present invention (claim 1), the active disk control device copies the storage information of the active disk device to the shared disk device, and the backup disk control device copies the storage information of the shared disk device to the shared disk device. Since the data is copied to a disk device, the processing capacity of the current central processing unit is not overwhelmed.
It becomes possible to copy the storage information of the current disk device to the spare disk device, and according to the present invention (claim 2), even when the current disk control device does not operate normally, the current central processing unit can control the spare disk. It becomes possible to extract the storage information of the current disk device through the device, and the reliability of the information processing system is greatly improved.

[Brief explanation of the drawing]

[Figure 1] Diagrams showing the principle of the present invention, where (a) is a diagram showing the principle of the present invention (Claim 1), and Figure (b) is a diagram showing the principle of the present invention (Claim 2).

[Fig. 2] A diagram showing an information processing system according to an embodiment of the present invention (Claim 1)

[Figure 3] A diagram showing an example of the disk control device in Figure 2.

[Fig. 4] A diagram showing an information processing system according to an embodiment of the present invention (Claim 3)

[Figure 5] A diagram showing an example of the bus switching device in Figure 4.

[Figure 6] Diagram showing an example of a conventional information processing system

[Explanation of symbols]

1 Central control unit (CC) 2 Main memory (MM) 3 Channel control unit (CHC) 4 Channel bus 5 Disk control unit (DKC) 6 Disk unit (DK) 7 Interlacing line 8 Shared disk unit (CDK) 9 Bus switching Device (BSW) 10 Central processing unit 51 Processing unit (MPU) 52 Read-only memory (ROM) 53 Write/read memo (RAM) 54 Channel bus connection control unit (CBIF) 55, 5
6, 58 Disk connection control unit (DIF) 57 Buffer memory (BFM) 59 Failure detection unit 91 Relay control unit 92 Relay (RY) 93 Relay contact (RY) 100 First connection means 200 Second connection means 300 Copying means 400 Third connection means 500 Switching means 521 Copying section (CPY) 551 Drive section (DV) 552 Receiving section (RV) 553 Control section (CTL)

Claims (5)

[Claims] Claim 1: Consisting of a dual system each comprising a central processing unit (10), a channel control device (3), a disk control device (5), and a disk device (6), one system serving as the active system. In an information processing system that is in operation and the other system is on standby as a standby system, a shared disk device (8) that can be written to and read from each system is provided,
Each disk control device (5) is provided with a first connection means (100) for controlling the disk device (6) in the same system independently from the disk device (6) in another system and the shared disk device (8). ), a second connection means (200) for controlling the shared disk device (8) independently of each disk device (6), and a disk device (200) controlled by the first connection means (100). 6) and the second connection means (
200) and a shared disk device (8) controlled by a shared disk device (8), a duplex disk control system is provided with a copying means (300) for copying information stored in one device to the other device. 2. Consisting of a dual system each comprising a central processing unit (10), a channel control device (3), a disk control device (5), and a disk device (6), one system serving as the active system. In an information processing system in which the other system is in operation and the other system is on standby as a standby system, each disk control device (5) is configured to control the disk device (6) in the same system independently from the disk device (6) in another system. and a third connection means (400) for controlling a disk device (6) in another system independently of the disk device (6) in the same system, Connecting a disk device (6) in the same system to the first connection means (100) of the disk control device (5) in the same system while the disk control device (5) in the system is operating normally; If the disk control device (5) in the same system does not operate normally, the disk device (6) in the same system is connected to the third connecting means (5) of the disk control device (5) in the other system. A duplex disk control system characterized in that each system is provided with a switching means (500) for switching connection to (400). 3. The normality of each disk control device (5) is monitored within each disk control device (5), and the monitoring result is notified to the switching means (500) in the same system. 3. The duplex disk control method according to claim 2. 4. The duplex disk control system according to claim 2, wherein the normality of each disk control device (5) is monitored by the switching means (500) within the same system. 5. A power source for each disk control device (5) is provided for each system independently of a power source for the disk device (6) and switching means (500). Duplex disk control method.