JP4716492B2 - Redundant configuration system and method for second computer system to immediately recover when first computer system fails - Google Patents

Description

  The present invention relates to a technique for performing high-speed recovery in a system having a redundant configuration, and particularly to a technique for recovering in a very short time.

In order to improve the reliability of the computer system, in addition to the primary system that operates normally, a secondary system that backs up the primary system in the event of a failure of the primary system is built, and a redundant system is constructed. ing. In this case, in order to immediately start processing in the secondary system when a failure occurs in the primary system, it is necessary to transfer the processing result data of the primary system to the secondary system. Therefore, in many cases, the main system and the sub system share one database, or the data of the main system is transferred to the sub system.
Japanese Unexamined Patent Publication No. 7-239799

  For example, even in the case of a system that processes a large amount of data at high speed and does not allow a temporary system down, the system is duplicated. In such a system, when a failure occurs in the primary system, the system must be switched instantaneously (for example, about 1 second) to the secondary system.

  However, in the conventional data transfer method, it takes time for the secondary system to be able to use the shared database, and switching to the secondary system cannot be performed instantaneously.

  Therefore, an object of the present invention is to immediately switch from a primary system to a secondary system in a system having a redundant configuration.

  A system according to an embodiment of the present invention is a redundant configuration system having a first computer system and a second computer system for backing up the first computer system. The first computer system includes a first receiving unit that receives an input message, a processing unit that executes a predetermined process based on the input message received by the first receiving unit, and a process performed by the processing unit. First execution result storage means for storing execution result information; and transfer means for transferring execution result information stored in the first execution result storage means to the second computer system. The second computer system includes second receiving means for receiving the same input message as the first computer system, and second input message storage means for storing the input message received by the second receiving means. A second execution result storage means for storing execution result information transferred from the first computer system; and an input stored in the second input message storage means in the event of a failure of the first computer system. Recovery means for executing the same processing as the first computer system based on the message and storing the execution result information in the second execution result storage means.

  In a preferred embodiment, in the first computer system, the first receiving means sequentially receives a plurality of input messages including respective identification information. The execution result information stored in the first execution result storage means associates the identification information of the input message with the execution result of the predetermined process executed based on the input message of the identification information. It may be. And the said 2nd computer system is memorize | stored in the said 2nd input message memory | storage means based on the identification information of the input message contained in the execution result information transferred from the said 1st computer system A means for updating information may be further provided.

  In a preferred embodiment, the first computer system may further comprise first input message storage means for storing an input message received by the first reception means. The second computer system may further comprise means for notifying the first computer of identification information of an input message that has been stored in the second input message storage means. Then, the processing means of the first computer system acquires an input message from the first input message storage means based on the identification information of the input message notified from the second computer, and A process may be executed.

  In a preferred embodiment, the processing means of the first computer system may execute a validation process for input data included in an input message.

  In a preferred embodiment, the transfer means of the first computer system may transfer the execution result information asynchronously with the processing means of the first computer system.

  Hereinafter, a data processing system having a redundant configuration according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a configuration diagram of a data processing system according to the present embodiment.

  As shown in the figure, this system includes business servers 1 and 2 that receive an input message 7 transmitted from a data transmission device 5 and execute a predetermined process based on the input message 7. The business server employs a redundant configuration having a primary business server 1 and a secondary business server 2.

  Each of the primary business server 1 and the secondary business server 2 is configured by, for example, a general-purpose computer system, and each component or function in each device 1 and 2 described below executes, for example, a computer program It is realized by doing.

  The main business server 1 includes a reception processing unit 10 that receives the input message 7 transmitted from the data transmission device 5, a response transmission unit 11 that notifies the data transmission device 5 of the processing result of the input message 7, and the received input An entry queue 12 for storing messages 7, a business processing unit 13 for executing predetermined business processing based on the input message 7, a business data storage unit 14 for storing business data generated by the business processing unit 13, and a business A data transfer unit 15 that transfers data to the subordinate business server 2 and an external interface unit 16 that provides business data to an external device are included.

  On the other hand, since the secondary business server 2 is a backup server when a failure occurs in the primary business server 1, it has a configuration similar to that of the primary business server 1 and a configuration unique to the secondary system. That is, the secondary business server 2 includes the reception processing unit 20 that receives the input message 7, the response transmission unit 21 that notifies the data transmission device 5 of the processing result of the input message 7 in the business processing unit 23, and the received input An entry queue 22 for storing messages 7, a business processing unit 23 for executing predetermined business processing based on the input message 7, a business data storage unit 24 for storing business data generated by the business processing unit 23, Data receiving unit 25 that receives business data transferred from the system business server 1, an external interface unit 26 for providing business data to an external device, and an entry that manages the input message 7 in the entry queue 22 And a queue management unit 27.

  First, normal processing of the primary and secondary business servers 1 and 2 having the above-described configuration will be described.

  A plurality of input messages 7 are sequentially transmitted from the data transmission device 5 to both the primary business server 1 and the secondary business server 2. The reception processing unit 10 of the primary business server 1 and the reception processing unit 20 of the secondary business server 2 receive the input message 7 and store them sequentially in the entry queues 12 and 22 respectively. That is, both the primary and secondary business servers 1 and 2 receive the same input message 7. Here, each input message 7 includes a message ID for uniquely identifying each input message.

  FIG. 2 shows the data structure of the entry queues 12 and 22. That is, in the entry queues 12 and 22, an input message 51 and a read flag 52 indicating whether or not the input message has been read are stored in association with each other.

  In the primary business server 1, the business processing unit 13 reads the input messages 51 stored in the entry queue 12 one by one, and sets a flag indicating that the data has been read in the read flag 52 corresponding to the read input message 51. . Then, the job processing unit 13 executes predetermined job processing on the read input message 51. The business process performed here is, for example, validation for checking whether the attribute and the number of digits of the input data included in the input message 51 are within a predetermined range, or various business processes based on the input data, or A combination of these may be used.

  When the business process for one input message 51 is completed, the business processing unit 13 stores the business data generated by the business process in the business data storage unit 14. At this time, the business process unit 13 notifies the response transmission unit 11 of the message ID included in the input message 51 for which the business process has been completed.

  Upon receiving the message ID notification from the business processing unit 13, the response transmission unit 11 notifies the data transmission device 5 of a response message indicating that the processing of the message ID has been normally completed.

  Note that the reception processing units 10 and 20 and the business processing units 13 and 23 may operate in different threads. Thereby, the reception processing units 10 and 20 and the business processing units 13 and 23 can operate asynchronously with each other.

  Next, FIG. 3 shows the data structure of the business data storage units 14 and 24. That is, in the business data storage units 14 and 24, the message ID 61 and the business data 62 after execution of business processing are stored in association with each other.

  The business processing unit 23 of the secondary business server 2 can execute the same processing as the business processing unit 13 of the primary business server 1. However, when the main business server 1 is operating normally, the business processing unit 23 does not execute processing. Data is stored in the business data storage unit 24 of the secondary business server 2 as follows.

  The data transfer unit 15 of the primary business server 1 acquires the message ID 61 and the business data 62 stored in the business data storage unit 14 and transfers them to the secondary business server 2. This is because the primary business server 1 and the secondary business server 2 are configured not to share the business data storage units 14 and 24, so the data stored in the business data storage units 14 and 24 must be matched. Because there is. This data transfer processing can be performed asynchronously with the processing of the business processing unit 13.

  In this embodiment, the primary business server 1 and the secondary business server 2 do not share the business data storage units 14 and 24, and data transfer from the primary business server 1 to the secondary business server 2 is also asynchronous. Therefore, the business processing unit 13 of the primary business server 1 can execute the processing without being aware of the presence of the secondary business server 2, and the processing speed can be increased.

  In the secondary business server 2, the transferred business data is received by the data receiving unit 25 and stored in the business data storage unit 24. Further, the message ID 61 of the data received by the data reception unit 25 is notified to the entry queue management unit 27.

  The entry queue management unit 27 updates the information stored in the entry queue 22 based on the message ID 61 included in the transferred business data. For example, the entry queue management unit 27 deletes the input message having the same message ID as the message ID 61 of the data transferred from the primary business server 1 from the entry queue 22 or indicates that the read flag 52 has been read out. Set a flag to indicate. Thereby, it is possible to determine an input message in which no data is stored in the business data storage unit 24 during a recovery process described later.

  The external interface unit 16 of the primary business server 1 allows an external system (not shown) to refer to the business data 62 stored in the business data storage unit 14.

  As with the external interface unit 16 of the primary business server 1, the external interface unit 26 of the secondary business server 2 makes the external system refer to the business data storage unit 24. Reference may be restricted when the primary business server 1 is operating normally.

  The business data is accumulated in the business data storage unit 14 by the processing when the main business server 1 is operating normally as described above. On the other hand, in the secondary business server 2, business data generated and transferred by the primary business server 1 is stored in the business data storage unit 24 and input data that has not yet been transferred from the primary business server 1. The input message 51 corresponding to is left in the entry queue 22. Such a state continues while the main business server 1 is operating normally.

  When a failure occurs in the primary business server 1, the secondary business server 2 performs the following processing to recover.

  That is, when a failure occurs in the primary business server 1, the secondary business server 2 is notified to that effect. When receiving the notification, the subordinate business server 2 starts the processing by the business processing unit 23. For example, the business processing unit 23 acquires an input message from the entry queue 22, executes the same input processing as the business processing unit 13, and stores the processed data in the business data storage unit 24.

  Here, since it is possible to determine whether all the input messages remaining in the entry queue 22 are unprocessed or unprocessed, the processed data is processed again. There is no loss of time.

  Further, the secondary business server 2 does not share the business data storage unit 24 with the primary business server 1. For this reason, the business data storage unit 24 can be realized not on an external storage device such as a hard disk but on a memory having a high access speed such as a RAM in the secondary business server 2. Accordingly, the timing at which the business processing unit 23 starts the process for recovery is earlier, and the recovery process can be started in a very short time from the occurrence of a failure in the primary business server 1.

  The above-described embodiments of the present invention are examples for explaining the present invention, and are not intended to limit the scope of the present invention only to those embodiments. Those skilled in the art can implement the present invention in various other modes without departing from the gist of the present invention.

  For example, the business data storage units 14 and 24 may be composed of one table or a plurality of tables. In the case of a plurality of tables, the update of these tables is handled as one transaction.

  When the reception processing unit 20 of the secondary business server 2 completes the storage of the received input message 7 in the entry queue 22, it generates a confirmation message including the message ID of the input message 7, and the primary business server 1 may be notified. In this case, when the primary business server 1 receives the confirmation message from the secondary business server 2, the business processing unit 13 reads out the input message 51 of the message ID included in the confirmation message from the entry queue 12 and determines the predetermined message. Execute the business process.

It is a block diagram of the data processing system which concerns on one Embodiment of this invention. It is a figure which shows an example of the data structure of the entry queues 12 and 22. 4 is a diagram illustrating an example of a data structure of business data storage units 14 and 24. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Primary business server 2 Sub business server 5 Data transmission device 7 Input message 10, 20 Reception processing unit 11, 21 Response transmission unit 12, 22 Entry queue 13, 23 Business processing unit 14, 24 Business data storage unit 15 Data transfer Units 16, 26 external interface unit 27 entry queue management unit

Claims (3)

A redundant system having a first server computer and a second server computer for backing up the first server computer,
The first server computer is
First receiving means for receiving an input message including identification information;
First processing means for executing predetermined business processing based on an input message received by the first receiving means;
First execution result storage means for storing execution result information of the predetermined job process executed by the first processing means in association with identification information of the input message;
Transfer means for transferring the execution result information stored in the first execution result storage means together with the identification information of the associated input message to the second computer system,
The second server computer is
Second receiving means for receiving the same input message as the first server computer;
Second input message storage means for storing an input message received by the second receiving means ;
Second execution result storage means for storing the execution result information transferred from the first server computer and the identification information of the input message in association with each other;
The input message stored in the second input message storage means corresponds to the processed message in which the execution result information of the corresponding predetermined business process is stored in the second execution result storage means. Determination means for determining, based on the identification information, an unprocessed message in which execution result information of the predetermined business process is not stored ;
Second processing means for executing the same business process as the first processing means based on the input message stored in the second input message storage means, wherein the first server computer is normal If the first server computer fails, the same business process as the first processing unit is executed based on the input message determined as an unprocessed message by the determination unit. redundant system comprising, and stores the execution result information in association with the identification information of the unprocessed message to the second execution result storage means, and said second processing means.   The redundant configuration system according to claim 1, wherein the transfer unit of the first server computer transfers the execution result information asynchronously with the processing unit of the first server computer. A method in which a second server computer immediately recovers when a failure occurs in a first server computer,
The first server computer and the second server computer receiving an input message including identification information;
The second server computer storing the received input message in the input message storage means of the second server computer;
The first server computer executing predetermined business processing based on the received input message;
Storing the execution result information of the business process in association with the identification information of the input message in the first execution result storage unit of the first server computer;
Transferring the execution result information stored in the first execution result storage means from the first server computer to the second server computer together with the identification information of the associated input message;
Storing the transferred execution result information and the identification information of the input message in association with each other in a second execution result storage unit of the second server computer;
The second server computer has processed the input message stored in the input message storage means, and the execution result information of the corresponding predetermined business process is stored in the second execution result storage means Determining, based on the identification information, a message and an unprocessed message in which execution result information of the corresponding predetermined business process is not stored;
The second server computer is not carried out when the first server computer is operating normally, when the first server computer failure, based on the input message is determined as unprocessed message by pre Symbol decision And executing the same business process as the first server computer and storing the execution result information in the second execution result storage means in association with the identification information of the unprocessed message .

Images