JP2010009293A - Computer system and system switching method - Google Patents

Abstract

It is possible to perform high-speed system switching of a computer system composed of an active computer and a standby computer.
An active computer and a standby computer share a single disk that stores an OS, and a monitoring computer that monitors the status of these computers and controls system switching is used. It is connected to the primary and standby computers. When the system is started up, the monitoring computer 103 turns on the power of the active and standby computers, causes the active and standby computers to perform hardware initialization and self-test, and then the active computer When the failure of the active computer is detected, the standby computer is made to take over the processing performed by the active computer.
[Selection] Figure 1

Description

  The present invention relates to a computer system and a system switching method, and more particularly to a computer system and a system switching method in which a plurality of computers are multiplexed by an active system and a standby system to enable system switching.

  In the modern society, it is required to provide various services through the Internet etc. without using a computer system for 24 hours 365 days a day. Therefore, a highly reliable computer system is required that can recover from a failure and continue providing services in the shortest possible time even when a failure occurs.

  One method for realizing a highly reliable computer system is to use a plurality of computers in a multiplexed manner. In this method, a computer system is configured by providing a standby computer that takes over the processing of the active computer. As a result, even when a failure occurs in the active computer, the system can be switched to the standby computer, and the standby computer can continue processing.

  A method called cold standby is known as a method for switching the system. In this system switching method called cold standby, a computer system is configured by an active computer and a standby computer that share a system disk storing an OS and applications necessary for providing a service, and there is a problem with the active computer. When this occurs, the standby computer is started and the processing is continued by switching the system. Although the reliability of the computer system can be improved by using the system switching by the cold standby as described above, this method starts the standby system after the failure occurs. However, it takes about 30 minutes to 1 hour to provide the service.

  As a conventional technique that enables a computer to be started at high speed, there is a high-speed boot technique. This high-speed boot technology makes it possible to realize a high-speed startup by omitting the hardware self-test performed at the time of startup of the computer. However, since this technology cannot start the OS or application when there is a failure in the computer hardware to be started, the hardware is reset and the hardware self-test is performed. Since the process of separating the part has to be performed, high-speed activation cannot be performed.

For example, a technique described in Patent Document 1 is known as a conventional technique related to a high-speed boot technique that is a technique for starting the computer at a high speed.
Japanese Unexamined Patent Publication No. 2007-4787

  Computer processing in which a computer system is configured by the active computer and the standby computer sharing the system disk storing the OS and applications necessary for providing the service described above, and the system can be switched by cold standby. In order to start the standby computer after a failure occurs in the active computer, it takes about 30 minutes to 1 hour until the standby computer completes startup and provides the service. Meanwhile, there is a problem that the service is stopped.

  The above-described problem can be solved to some extent by applying the above-described high-speed boot technique to a standby computer. However, even if fast boot technology is applied, the computer system in that case cannot start the OS or application if there is a failure in the hardware of the standby computer. Since it is necessary to perform a test, isolate the faulty part, and start up, the standby system completes the start-up and the time required to provide the service cannot be shortened.

  In view of the above-described points, an object of the present invention is to provide a computer system and a system switching method that can perform system switching at high speed in a computer system composed of an active computer and a standby computer. There is.

  According to the present invention, the object is to provide a computer system in which the active computer and the standby computer share the disk storing the operating system, and the standby computer performs hardware initialization and startup when the computer system is started up. This is achieved by waiting in a state where a self-test has been performed and taking over the processing performed by the active computer when the failure of the active computer is notified.

  The object is to monitor the status of the active computer and the standby computer and control the system switching in a computer system in which the active computer and the standby computer share the disk storing the operating system. A monitoring computer is connected to the active computer and the standby computer, and when the computer system is started up, the monitoring computer turns on the power of the active computer and the standby computer, and Causing the active computer and the standby computer to perform hardware initialization and self-test; then, causing the active computer to execute a process of providing a service; causing the standby computer to wait in that state; and the active computer When a failure is detected, Is accomplished by to take over the active computer is performing processing to the machine system computer.

  According to the present invention, even when there is a failure that can be activated in the hardware of the standby computer, it becomes possible to quickly bring the standby computer into an operating state and perform system switching at high speed. As a result, the processing performed on the active computer can be promptly continued on the standby computer.

  Embodiments of a computer system and a system switching method according to the present invention will be described below in detail with reference to the drawings.

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

  The computer system according to the embodiment of the present invention shown in FIG. 1 includes an active computer 101, a standby computer 102, and a monitoring computer 103, a network control device 114 provided in the active and standby computers 101 and 102, and monitoring. It is configured to be connected via a network control device 126 provided in the system computer 103.

  The active computer 101 includes a CPU 111, a memory 105, a disk control device 112, a power supply control device 113, and a network control device 114. Although the internal configuration of the standby computer 102 is not shown in FIG. 1, it is configured in exactly the same way as the active computer 101. The active computer 101 and the standby computer 102 share one disk 104. The disk 104 stores an OS boot loader 115, an OS 116, and an application 117.

  When the active computer 101 or the standby computer 102 is activated, the contents of the disk 104 are read into the memory 105. In addition, firmware stored in a ROM (not shown) in the computer is read into the memory 105. As a result, the application 106, the OS 108, and the firmware 109 are read into the memory 105. The application 106 includes a heartbeat instruction sequence 107. In addition, the firmware 109 includes an activation restart instruction sequence 110.

  The monitoring computer 103 includes a CPU 125, a memory 118, a network control device 126, a disk control device 127, a disk 128, and a console control device 129. When the monitoring computer 103 is activated, the application program 119 and the OS 123 are read into the memory 118 from the disk 128 and the firmware 124 is read from the ROM (not shown). The application 119 includes an active standby system management table 120, a start command sequence 121, and a return command sequence 122.

  In the embodiment of the present invention described above, a computer system is configured by including one active computer, one standby computer, and one monitoring computer, but the present invention is provided for a plurality of active computers. A computer system can be configured by providing one standby computer.

  FIG. 2 is a diagram showing a configuration example of the active standby system management table 120 included in the application 119 of the monitoring computer 103. The management table 120 includes a column of a computer identifier 202, a column of an active / standby system 203 indicating whether it is an active system or a standby system, and a column of a status 204 indicating the state of a computer.

  In the column of the computer identifier 202, information on the IP addresses 205 and 206 of the computers constituting the computer system according to the embodiment of the present invention is stored. The active / standby system 203 column indicating whether the current system is the active system or the standby system has an active system 207 that indicates whether the computers are the active system or the standby system in correspondence with the rows of the IP addresses 205 and 206 of the computers. Information on the standby system 208 is stored. Further, the status column 204 stores information on three types of statuses of stop, standby, and operation as the status of the computer that is the active or standby system. In the illustrated example, stops 209 and 210 are stored in both.

  FIG. 3 is a flowchart for explaining the processing operation of the heartbeat instruction sequence 107 in the application 106 of the active computer 101 and standby computer 102. Next, this will be described.

(1) The heartbeat command sequence determines whether or not a heartbeat confirmation packet sent from the monitoring computer 103 has been received. If not received, it is determined whether or not a heartbeat confirmation packet has been received until reception. Is repeated (step 302).

(2) If a heartbeat confirmation packet is received in the determination in step 302, a heartbeat response packet is transmitted to the monitoring computer 103 (step 303).

  The heartbeat command sequence 107 can inform the monitoring computer 103 that the own computer is alive by repeating the series of processes described above.

  FIG. 4 is a flowchart for explaining the processing operation of the firmware 109 of each of the active and standby computers. Next, this will be described.

(1) When the power supply is turned on, the active computer and the standby computer start up the firmware 109. The firmware 109 executes a hardware initialization process and a self-test. The existing hardware is disconnected (step 402).

(2) Next, the firmware 109 calls the boot restart instruction sequence 110, and then passes the processing to the OS boot loader 115, and ends the processing here (steps 403 to 405).

  As described above, the OS boot loader 115 loads the OS 116 in the disk 104 into the memory 105 to make the active and standby computers operable.

  FIG. 5 is a flowchart for explaining the processing operation of the activation / resumption instruction sequence 110 in the firmware 109 of each of the active and standby computers. Next, this will be explained.

(1) The start / resume instruction sequence 110 is a process of step 402 described with reference to FIG. 4. When the hardware initialization process and the self-test are executed by turning on the computer, the self-test completion packet is transmitted to the monitoring computer. 103 is transmitted. Since each of the active and standby computers 101 and 102 does not have information for specifying the monitoring computer 103, the self-test completion packet is transmitted by broadcast (step 502).

(2) Thereafter, the activation resumption instruction sequence 110 determines whether or not an activation resumption packet has been received from the monitoring computer 103, and when received, resumes the activation process. That is, the processing from step 404 in FIG. 4 is started. If the activation restart packet cannot be received, the activation of the own computer remains suspended in the self-test completed state until it is received (step 503).

  6A, 6B, and 6C are flowcharts for explaining the processing operation of the activation instruction sequence 121 possessed by the application 119 of the monitoring computer 103, which will be described next. 6A, 6B, and 6C show a series of processing operations, the following description will be described as a series of processes, and will also be described as a process executed by the monitoring computer.

(1) The monitoring computer 103 transmits a packet for turning on the power of each computer to the active and standby computers 101 and 102 (step 602).

(2) Each of the active and standby computers 101 and 102 that has received the packet for turning on the power turns on its own computer and performs the processing described with reference to FIG. Since a completion packet is transmitted after completion, it is determined whether or not a self-test completion packet has been received from each of the computers 101 and 102 of the active system and the standby system. The states of the active system 209 and the standby system 210 in the queue are changed from the initial state stop to the standby state (steps 603 and 604).

(3) If it is determined in step 603 that the self-test completion packet has not been received, the operation of the monitoring computer 103 is suspended until it is received.

(4) After changing the status 204 of the management table 120, the monitoring computer 103 transmits an activation restart packet and a heartbeat confirmation packet to the active computer 101 (steps 605 and 606).

(5) It is determined whether or not a heartbeat response packet has been received from the active computer 101. If a heartbeat response packet is received, it is considered that the active computer 101 has been activated and the status 209 of the active system in the management table 120 Is changed from standby to operation (steps 607 and 608).

(6) If the heartbeat response packet is not received in the determination in step 607, the processing returns to step 606, the heartbeat confirmation packet is transmitted again, and the processing is repeated until the heartbeat response packet is received.

(7) After changing the state 204 of the management table 120 in the process of step 608, a timer is set and a heartbeat confirmation packet is transmitted to the active computer 101 (steps 609 and 610).

(8) It is determined whether or not a heartbeat response packet has been received from the active computer 101. If a heartbeat response packet is received, the process returns from step 609 to continue the process from the process of setting the timer again. (Step 611).

(9) If the heartbeat response packet is not received in the determination in step 611, the timer is advanced to determine whether the time specified by the timer has elapsed. If the specified time has not elapsed, from step 611 Returning to the process, the process from the determination process of whether or not the heartbeat response packet has been received is continued (steps 612 and 613).

(10) If it is determined in step 613 that the timer has elapsed for a certain period of time, it is regarded as a heartbeat timeout due to a failure of the active computer 101, the active computer 101 is turned off, and the active system in the status 204 of the management table 120 is turned off. The state 209 is changed from operation to stop (steps 614 and 615).

(11) Thereafter, the monitoring computer 103 transmits an activation restart packet and a heartbeat confirmation packet to the standby computer 102 (steps 616 and 617).

(12) It is determined whether or not a heartbeat response packet has been received from the standby computer 102. If a heartbeat response packet is received, it is considered that the standby computer 102 has been operated, and the standby state 210 in the management table 120 Is changed from standby to operation (steps 618 and 619).

(13) If the heartbeat response packet is not received in step 618, the process returns to step 617, the heartbeat confirmation packet is transmitted again, and the process is repeated until the heartbeat response packet is received.

(14) The monitoring computer 103 changes the standby system state 210 of the management table 120 from standby to active, and then the active system in the active / standby system 203 column indicating whether the management table 120 is active or standby. Display 207 is changed to the standby display, and the standby display 208 is changed to the active display (step 620).

(15) Next, the monitoring computer 103 displays a message in the management table 120 to replace the computer that has become the standby display via the console control device 129 (step 621).

(16) After the message is displayed in step 621, the maintenance staff shall promptly replace the computer, and after the replacement, the maintenance staff executes the return instruction sequence 122 in the application 119 of the monitoring computer 103. Shall. Thereby, the standby computer 102 and the active computer 101 can be used, and the processing from the timer setting can be continued by returning to the processing from step 609.

  FIG. 7 is a flowchart for explaining the processing operation of the return instruction sequence 122 possessed by the application 119 of the monitoring computer 103. Next, this will be explained.

(1) When the maintenance engineer starts up the return instruction sequence 122 after the replacement of the computer, the monitoring computer 103 is stopped due to a failure in the standby computer 102 (in the example of the embodiment of the present invention described). A packet for turning on the power is transmitted to the computer after replacing the computer (step 702).

(2) The standby computer 102 turns on the power according to an instruction from the monitoring computer 103 and transmits a self-test completion packet to the monitoring computer 103 when the hardware self-test is completed. Waits until the packet is received (step 703).

(3) When the monitoring computer 103 receives the self-test completion packet, the monitoring computer 103 executes the standby system state 209 in the column of the state 204 of the management table 120 (the return instruction sequence is executed after switching between the standby system and the active system). Since this is an instruction sequence, the location of the information indicating the state of the standby system here corresponds to 209) is changed from stop to standby, and the processing here ends (step 704).

  FIG. 8 is a sequence chart showing the processing in the embodiment of the present invention described as the processing of the monitoring computer shown in FIGS. 6A to 6C and FIG. 7 as processing among the active, standby, and monitoring computers. Next, this will be described.

(1) The monitoring computer 103 instructs each of the active and standby computers 101 and 102 to turn on the power of each computer (steps 801 and 802).

(2) In response to an instruction from the monitoring computer 103 in steps 801 and 802, each of the active and standby computers 101 and 102 turns on its own computer, initializes the hardware, and performs a self-test. Are executed (steps 803 and 804).

(3) After completion of the hardware self-test, each of the active and standby computers 101 and 102 transmits a completion packet to the monitoring computer 103 (steps 805 and 806).

(4) When the monitoring computer 103 receives the self-test completion packet from each of the active and standby computers 101 and 102, the monitoring computer 103 transmits a start restart packet to the active computer 101, and the standby computer 102 ,do nothing. As a result, the standby computer 102 is on standby from this point (step 807).

(5) The active computer 101 that has received the activation restart packet initializes the OS and initializes the application, starts the service, and enters the service providing state (steps 808 to 810).

(6) While the active computer 101 is providing services, the monitoring computer 103 transmits a heartbeat confirmation packet to the active computer 101 and receives a heartbeat response packet from the active computer 101 at regular intervals. Repeatingly, it is confirmed that the active computer 101 normally provides a service (steps 811 and 812).

(7) If the active computer 101 fails during service provision and stops the service, the active computer 101 cannot transmit a heartbeat response packet to the heartbeat confirmation packet from the monitoring computer 103. As a result, the monitoring computer 103 detects a failure of the active computer 101 (steps 813 to 816).

(8) The monitoring computer 103 that has detected the failure of the active computer 101 instructs the active computer 101 to turn off the power, and transmits an activation restart packet to the standby computer 102 (steps 817 and 818). ).

(9) The standby computer 102 that has received the activation restart packet initializes the OS and initializes the application, starts the service as the active computer, and enters the service providing state (steps 819 to 821). .

(10) While the standby computer 102 is providing the service, the monitoring computer 103 transmits a heartbeat confirmation packet to the standby computer 102 and receives a heartbeat response packet from the standby computer 102 at regular intervals. Repeatingly, it is confirmed that the standby computer 102 normally provides a service (steps 822 and 823).

(11) On the other hand, the active computer 101 that has failed and stopped the service is dealt with by replacement of the computer by the maintenance personnel, and then the restoration instruction sequence of the monitoring computer 103 is started by the maintenance personnel (step 824). 825).

(12) When the return instruction sequence is activated, the monitoring computer 103 instructs the stopped active computer 101 to turn on the power (step 826).

(13) In response to the instruction from the monitoring computer 103, the active computer 101 turns on its own computer, executes hardware initialization and self-test, and completes after completing the hardware self-test. The packet is transmitted to the monitoring computer 103 to enter a standby state (steps 827 and 828).

  Each process in the above-described embodiment of the present invention is configured by a program and can be executed by a CPU included in the present invention. These programs are stored in a recording medium such as an FD, CDROM, or DVD. It can be provided and can be provided by digital information via a network.

It is a block diagram which shows the structure of the computer system by one Embodiment of this invention. It is a figure which shows the structural example of the active system standby system management table contained in the application of a monitoring system computer. It is a flowchart explaining the processing operation | movement of the heartbeat command sequence in the application of an active computer and a standby computer. It is a flowchart explaining the processing operation of the firmware of each computer of an active system and a standby system. It is a flowchart explaining the processing operation | movement of the starting resumption command sequence in the firmware of each computer of an active system and a standby system. It is a flowchart (the 1) explaining the processing operation | movement of the starting command sequence which the application of a monitoring system computer has. It is a flowchart (the 2) explaining the processing operation | movement of the starting command sequence which the application of a monitoring system computer has. It is a flowchart (the 3) explaining the processing operation | movement of the starting command sequence which the application of a monitoring system computer has. It is a flowchart explaining the processing operation | movement of the return instruction sequence which the application of a monitoring computer has. FIG. 8 is a sequence chart illustrating the processing described with reference to FIGS. 6A to 6C and FIG. 7 as processing between the active, standby, and monitoring computers.

Explanation of symbols

101 Active computer 102 Standby computer 103 Monitoring computer 104, 128 Disk 105, 118 Memory 106, 117, 119 Application 107 Heartbeat instruction sequence 108, 116, 123 OS
109, 124 Firmware 110 Activation restart instruction sequence 111, 125 CPU
112, 127 Disk control device 113 Power supply control device 114, 126 Network control device 115 OS boot loader 120 Active standby system management table 121 Start command sequence 122 Return command sequence 129 Console control device

Claims (3)

In a computer system in which an active computer and a standby computer share a disk storing an operating system,
The standby computer waits in a state where hardware initialization and self-test are performed when the computer system is started up, and when the failure of the active computer is notified, the processing performed by the active computer A computer system characterized by taking over. In a computer system in which an active computer and a standby computer share a disk storing an operating system,
A monitoring computer that monitors the status of the active computer and the standby computer and controls system switching is connected to the active computer and the standby computer;
When the computer system is started up, the monitoring computer turns on the active and standby computers to cause the active and standby computers to perform hardware initialization and self-test. The active computer was executed on the standby computer when the active computer was caused to execute a process of providing a service, the standby computer was in that state, and a failure of the active computer was detected. A computer system characterized in that processing can be taken over. In a system switching method in a computer system in which a working computer and a standby computer share a disk storing an operating system,
A monitoring computer that monitors the status of the active computer and the standby computer and controls system switching is connected to the active computer and the standby computer;
When the computer system is started up, the monitoring computer turns on the active and standby computers to cause the active and standby computers to perform hardware initialization and self-test. The active computer was executed on the standby computer when the active computer was caused to execute a process of providing a service, the standby computer was in that state, and a failure of the active computer was detected. A system switching method characterized in that the processing is taken over.

Images