JP2001290670A - Cluster system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem where it is necessary to manually switch processings to a reserve computer system, each time a fault occurs in the conventional cluster system. SOLUTION: This system is provided with a fault-detecting means 101 for detecting the fault of a magnetic storage device 130 in an active computer system 100, a data-transmitting means 103 for transmitting data in the main storage device 110 in the active computer system to a reserve computer system 200 when the fault is detected and a data receiving means 201 for receiving the transmitted data and storing them in a main storage device 210 located on the side of the reserve computer system. After the data are completely transmitted/received, the active computer system is stopped, an application is started in the reserve computer system and by executing processing, while referring to the data stored in the main storage device 210, the relevant application continuously performs processing in the active computer system.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a cluster system including an active computer and a spare computer, and more particularly to a takeover system for transferring data of a main storage device to a spare computer when a failure occurs.

[0002]

2. Description of the Related Art Conventionally, such a cluster system is one of the following.
It is composed of one or more computer devices and one spare computer device. If a failure occurs in any one of the computer devices, the spare computer device takes over the processing of the failed computer device to operate the system. It is carried out.

[0003]

However, in the above-described conventional cluster system, when a failure occurs in a computer device, the processing is manually switched to a spare computer device, and the data in the main storage device is transferred to the spare computer device. They did not send and automate the process takeover. Therefore, it is necessary to manually hand over the processing to the spare computer every time a failure occurs.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned conventional problems, and an object of the present invention is to transmit data in a main storage device to a spare computer device when a failure occurs and to automatically take over processing. It is to provide a simple cluster system.

[0005]

According to the present invention, in order to achieve the above object, in a cluster system comprising a working computer device and a spare computer device, means for detecting a failure in an auxiliary storage device of the working computer device; Means for transmitting data in the main memory of the working computer to the spare computer when a fault is detected by the fault detecting means; means for receiving the transmitted data and storing the data in the main memory of the spare computer. After the transmission and reception of the data is completed, the active computer device is stopped, and an application is started in the spare computer device, and the application executes a process with reference to the data stored in the main storage device Thus, the present invention is characterized in that the processing of the active computer device is continuously performed.

[0006]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an embodiment of the cluster system of the present invention. In FIG. 1, the cluster system includes an active computer device 100 and a standby computer device 200. When a failure occurs in the active computer device 100, the processing is taken over by the standby computer device 200. Although the number of the active computer 100 is one in FIG.
0 may be plural.

The active computer device 100 includes a magnetic storage device 130 serving as an auxiliary storage device, an input / output control unit 120 for controlling data input / output of the magnetic storage device 130, and a failure detecting means 101 for detecting a failure of the magnetic storage device 130. An emergency stop unit 102 for emergency stop of the active computer 100 when a failure occurs, a main storage device 110, and a data transmission unit 103 for transmitting data on the main storage device 110 to the spare computer device 200 when a failure occurs. Reference numeral 300 denotes an application program (hereinafter, abbreviated as an application) on the active computer device 100.

[0008] The spare computer device 200 includes a data receiving means 201 for receiving data transmitted from the active computer device 100, a main storage device 210, an application starting means 202, a magnetic storage device 230, and an input / output control unit 220.
It is composed of Note that 400 is a computer terminal connected to the cluster system, and 500 is an application on the computer terminal 400. Here, in the present embodiment,
The failure detection unit 101 detects a failure in the magnetic storage device 130, and assumes a failure in the magnetic storage device 130 as a failure in which the processing of the application 300 cannot be continued in the active computer device 100.

Next, a specific operation of the present embodiment will be described in detail with reference to the flowcharts of FIGS. First, FIG. 2 is a flowchart showing the failure detection processing of the failure detection means 101. In FIG. 2, when the system is started, the failure detection unit 101 reads a target device (in this case, the magnetic storage device 130) from a setting value file (not shown),
A set value such as a time interval to be checked is acquired (step A1). The failure detection means 101 performs failure detection by test I / O based on the setting values obtained during operation of the system.

That is, the test I / O is
Is issued (step A2), and it is determined whether the data is normal based on the reply information to the test I / O from the magnetic storage device 130 (step A3). At this time, if it is normal,
After the failure detection means 101 has been stopped for a certain period of time (step A
4) Return to step A2 again, issue a test I / O, and determine whether or not it is normal (step A3). Less than,
The processing of steps A2 to A4 is repeated, and a test I / O is periodically issued to monitor whether the magnetic storage device 130 is normal. On the other hand, during the operation of the system, step A
When a disk failure of the magnetic storage device 130 occurs and a failure is detected in step 3, the failure detection unit 101 notifies the emergency stop unit 102 of the failure.

FIG. 3 is a flow chart showing the flow of processing of the emergency stop means 102 when a failure occurs. In FIG. 3, when the emergency stop unit 102 is notified of the occurrence of a failure, first, in order to put the application 300 into a blocked state,
A closing process is performed (step B1). The closed state refers to a state in which a transaction request to the application 300 is not accepted. Further, the closing process refers to a process of changing a closed state table (not shown) to a closed state and waiting for the end of a running transaction.

Here, the application 500 of the computer terminal 400 checks whether or not the application 300 is in a closed state by referring to the closed state table before issuing a transaction request to the application 300, and determines whether a new transaction request is possible. Are you sure? Therefore, the application 500 determines that the application 300 is in the blocked state after performing the blocking process, and does not issue a new transaction request.

When the closing process is completed, the data transmitting means 1
Reference numeral 03 denotes the data stored in the main storage device 110 and the spare computer device 2
00 is performed (step B2). FIG. 4 is a flowchart showing the processing of the data transmission means 103. In FIG. 4, first, the data transmission unit 103 acquires setting values such as a memory identifier and the size of data to be transmitted from a setting value file (not shown) (step C).
1).

Next, the memory (main storage device 110) is attached using the obtained identifier (step C).
2), data receiving means 201 on the standby computer 200
A socket is created and a connection is established to perform communication using the TCP / IP protocol with the server (step C).
3). Further, the data transmission unit 103 is connected to the main storage device 110.
Is read (step C4), and the data is transmitted (step C5). In this case, the data transmission means 10
Numeral 3 reads data from the main storage device 110 by the size obtained in the setting value file from the address obtained by the attachment, and transmits the data to the spare computer device 200.

Next, the processing of the data receiving means 201 will be described with reference to FIG.
This will be described with reference to the flowchart of FIG. In FIG.
First, activation of the data receiving means 201 is performed by the basic software (O
This is performed at the start of S). First, a set value such as a memory identifier and data size is obtained from the set value file (step D1), and a memory for storing data is secured based on the set value. Attachment of the memory (main storage device 210) is performed (step D2). At this time, the memory identifier and the size of the data are the same as the setting values in the case of the data transmission unit 103 in the active computer device 100.

Subsequently, the data receiving means 201 transmits the TCP /
As a communication procedure by IP, processing is performed to prepare a socket, associate with a port, set a queue, and wait for a connection request, and prepare for communication (step D3). Thus, the connection request can be accepted at any time from the data transmission unit 103 (step D).
4). In this state, when there is a connection request from the data transmitting unit 103, the connection is established, the transmission of data from the data receiving unit 201 is waited, and the data is received (step D5). The received data is stored in the main storage device 2
10 is written to the address obtained by the attachment (step D6).

Referring back to FIG. When data transmission / reception is completed in this way, the emergency stop unit 102 performs an emergency stop process of the application 300 in step B3 of FIG. Next, an emergency stop of the basic software (OS) is performed (step B4), and the stop processing of the active computer device 100 is completed.

On the other hand, the application 300 is started on the spare computer 200, and the processing of the active computer 100 is continued. FIG. 6 shows the processing of the application starting means 202 at this time. In FIG. 6, the application starting unit 202 acquires the same memory identifier as the setting value of the data receiving unit 201 from the setting value file (step E1), and starts the application 300 (step E2). In this case, the application 300
Using the obtained memory identifier, the process is performed by referring to the data in the main storage device 210, whereby it is possible to continue the process from before the switching of the active computer device 100.

Next, another embodiment of the present invention will be described. In the present embodiment, the failure detection method of the failure detection unit 101 is different. Other configurations are the same as those of the embodiment of FIG. FIG. 7 is a flowchart showing the processing of the failure detection means 101 of the present embodiment. 7, first, the failure detection unit 101 issues a test I / O to the magnetic storage device 130 (step F1), and detects a failure based on whether the result of the test I / O is normal (step F).
2). If normal, after stopping for a certain time (step F
3) Return to step F1, issue a test I / O again, and determine whether or not it is normal (step F2).

As described above, the test I / O is periodically issued to monitor whether the magnetic storage device 130 is normal.
Here, the failure detection means 101 counts the number of times that it is determined to be abnormal in step F2 (step F4), and compares the count value with a preset threshold (step F4).
5). In this case, if an intermittent failure occurs due to the deterioration of the disk of the magnetic storage device 130, the test I / O becomes abnormal.
It returns to 1 and treats it as normal.

As described above, in the present embodiment, the number of times determined to be abnormal is counted, and the occurrence of a failure is detected when the count value exceeds the threshold value. Unnecessary spare computer device 20
Switching to zero can be prevented. Therefore, the time required for switching the computer device can be reduced, and the processing efficiency of the system can be improved. Further, in the present embodiment, by adjusting the threshold value, it is possible to actively switch the computer device or to reduce the frequency of the switching.

[0022]

As described above, the present invention has the following effects. (1) When a failure occurs in an auxiliary storage device required for processing of an application, data in the main storage device of the active computer device is transmitted to the spare computer device and stored in the main storage device. In, the processing of the active computer device can be performed subsequently. (2) Consistent data can be taken over by performing a closing process of not accepting a transaction request for an application before transmitting / receiving data. (3) Since data transmission and reception are performed when the computer device is switched, there is no need to perform extra communication during normal system operation. (4) Even if there are a plurality of active computer devices, the spare computer device needs only the memory capacity of one active computer device, so that the memory capacity does not increase.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an embodiment of a cluster system according to the present invention.

FIG. 2 is a flowchart illustrating a process of a failure detection unit in FIG. 1;

FIG. 3 is a flowchart showing a process of an emergency stop means of FIG. 1;

FIG. 4 is a flowchart showing a process of a data transmission unit of FIG. 1;

FIG. 5 is a flowchart illustrating a process of a data receiving unit in FIG. 1;

FIG. 6 is a flowchart showing a process of an application starting unit in FIG. 1;

FIG. 7 is a flowchart illustrating a process of a failure detection unit according to another embodiment of the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 active computer device 101 failure detection means 102 emergency stop means 103 data transmission means 110 main storage device 120 input / output control unit 130 magnetic storage device 200 spare computer device 201 data reception means 202 application activation means 210 main storage device 220 input / output control unit 230 magnetic storage device 300 application 400 computer terminal 500 application

Claims (4)

[Claims] In a cluster system comprising a working computer and a spare computer, means for detecting a fault in an auxiliary storage device of the working computer, and a means for detecting a fault in the working computer when the fault is detected by the fault detecting means. Means for transmitting data in the main storage device to the spare computer device, and means for receiving the transmitted data and storing it in the main storage device on the side of the spare computer device, and Stopping and starting an application in the spare computer device, the application refers to data stored in the main storage device and executes a process to continue the process of the active computer device. And the cluster system. 2. The apparatus according to claim 1, further comprising means for performing a closing process for disabling a transaction request for an application when the failure detecting means detects a failure in the auxiliary storage device. Cluster system. 3. The method according to claim 2, wherein the failure detection unit periodically issues a test I / O to the auxiliary storage device and detects a failure in the auxiliary storage device based on a reply result to the test I / O. The cluster system according to claim 1, wherein 4. The fault detecting means counts the number of abnormalities as a result of the test I / O, and detects a fault in the auxiliary storage device when the count value exceeds a predetermined value. The cluster system according to claim 3.