JPH04273553A - Computer system - Google Patents

Abstract

PURPOSE:To omit the erasion of data by operating a power supply by a spare power source when the power supply from a main power source is stopped, and saving the data to a nonvolatile storage means. CONSTITUTION:When a fault occurs in a main power source device 3, and the power supply to the main body of a computer 1 is stopped, a spare power source device 16 starts the operation, operates the power supply to the main body of the computer 1 instead of the main power source device 1, and at the same time, operates an interruption to a CPU 11. At that time, the CPU 11 recognizes the occurrence of the fault of the power source by this interruption, and saves the data which are difficult to be restored and reactivated once erased among the data in a buffer for a secondary storage device 2 of a main storage part 13, to a nonvolatile memory 15. Then, the main body of the computer 1 stops the operation.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer system equipped with a function capable of dealing with power failures and the like.

0002

[Prior Art] Generally, a computer system has various I/O
It has a data buffer for the interface on the main memory. Therefore, for example, even if there is a request to change data to a secondary storage device outside the computer system, it is first reflected in the data buffer in the main storage, and then the data change in the secondary storage device is executed. It is now possible to do so.

[0003] Therefore, even when a computer system is to be stopped, the data in the buffer on the main memory is written back to the secondary storage device, and the system is stopped only after the data changes are reflected in the secondary storage device. .

[0004] From this, if the system were to stop without the changed data on the I/O buffer in the main memory being reflected in the secondary storage, the CPU would not be able to rewrite the data in the secondary storage. Data may not actually be rewritten even though it was intended to be rewritten. As a result, when the system is started up again and the contents of the secondary storage device are attempted to be referenced, the CPU may refer to incorrect data before the data in the secondary storage device has been rewritten.

For these reasons, for example, if the main power of a computer system is suddenly cut off due to an accident and the computer stops operating, only the changed data will be stored in the main memory buffer. Data that has not been reflected may disappear in the main memory when the main power is turned off.

Conventionally, in order to cope with a power failure such as a main power outage, backup power supplies are provided in each of the computer main body and the secondary storage device, and when a power failure occurs,
Power is supplied from each backup power supply to the computer main body and secondary storage device for a predetermined period of time, data in the main memory that needs to be written back to the secondary storage device is written back to the secondary storage device, and then , a system that halts the system is being considered. If this kind of processing is realized,
Data changes on the main storage section will be correctly reflected on the secondary storage device side. However, with this method, while performing the above operations, power must be supplied to the computer main body and the secondary storage device from each backup power supply, which increases the capacity of the required backup power supply, making it less economical. There are drawbacks that are disadvantageous to

[0007] Another method is to prepare two sets of identical computer systems that operate on different power supply systems, and have these two sets of computer systems execute the exact same processing at the same time. A method is being considered in which even if a system failure occurs, another computer can take over the processing. This method has the advantage of being able to deal with failures of the computer itself other than power failures such as power outages, but has the disadvantage of doubling the cost.

[0008]

[Problem to be Solved by the Invention] In this way, a computer system that is equipped with a function that can cope with a conventional power supply failure,
If a backup power supply is provided that can back up the entire computer system including the secondary storage device, the backup power supply would need to have a large capacity, which would be economically disadvantageous. Even if a system was prepared, it had the disadvantage of doubling the cost.

The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a computer system that can appropriately deal with power supply failures and that also enables cost reduction.

0010

[Means for Solving the Problems] The computer system of the present invention temporarily stores processing results inside the computer main body in a data buffer on the main storage section using power supplied from the main power supply, and then writes them to an external storage means. was made,
A backup power supply that supplies power to the computer when the main power supply is cut off, and a non-volatile power supply that uses the power supplied from the backup power supply to save at least the data that must not be lost in the data buffer on the main memory. It is composed of storage means.

[0011]

[Operation] As a result, according to the present invention, if the power supply from the main power source to the computer main body is cut off for some reason, power is supplied to the computer main body from the backup power source instead, and at the same time the main memory is At least the data in the local data buffer that must not be lost can be saved to a nonvolatile storage means.

0012

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows a schematic configuration of the same embodiment. In the figure, 1 is the computer main body, and this computer main body 1
are the CPU 11, the system bus 12, the main memory section 13, and the I
It has an /O interface 14, a nonvolatile memory 15, and a backup power supply 16. A secondary storage device (for example, a magnetic disk device, a semiconductor disk device) 2 and a main power supply device 3 are connected to this computer main body 1 via an I/O interface 14, and other data (not shown) is input. A terminal for output is connected.

In this case, the nonvolatile memory 15 of the computer main body 1 is, for example, an EEPROM, and the CPU 11
It is now readable and writable. The non-volatile memory 15 also has an area for storing a variable S indicating whether or not data is currently being saved; S=0 if the data is not being saved; S
= 1.

For example, a storage battery is used as the backup power supply device 16. The standby power supply device 16 is always charged by the main power supply device 3 and supplies power to the computer main body 1 when the power supply from the main power supply device 3 is cut off. In addition, this backup power supply device 16 supplies power to the computer main body 1 and at the same time supplies power to the computer main body 1.
It is also configured to issue an interrupt to U11. Then, due to this interrupt, the CPU 11 causes the main power supply unit 3 to
I am trying to know when the power supply has been cut off.

Next, the operation of the embodiment configured as above will be explained.

Now, if a failure occurs in the main power supply device 3 and the power supply to the computer main body 1 is stopped, the power supply shown in FIG.
The flowchart shown in is executed. In this case, when the main power supply 3 stops supplying power, the standby power supply 16 starts operating, supplies power to the computer main body 1 instead of the main power supply 3, and at the same time interrupts the CPU 11 (step A1 ).

[0018] When the CPU 11 learns of the occurrence of a power failure through this interrupt, it non-volatilely stores data in the buffer for the secondary storage device 2 of the main storage unit 13 that, if lost, would impede recovery and restart. The data is saved in the data memory 15 (step A2). In this case, the nonvolatile memory 15 has the variable S initialized to "0" in advance, and is set to S=1 by data saving (step A3). Then, the computer main body 1 stops operating.

Next, when the failure of the main power supply unit 3 is recovered and the system is started up, the flowchart shown in FIG. 3 is executed. In this case, the initialization program is started and
Refer to variable S in non-volatile memory 15 (step B1
). Here, if the data has not been saved in the nonvolatile memory 15 and S=0, it is determined as NO, and the OS is
(Step B2). On the other hand, if the data has been saved in the non-volatile memory 15 and S=1, it is determined as YES, and the data in the non-volatile memory 15 is written back to the secondary storage device 2 (step B3). The variable S in 15 is set to "0" (step B4), and then the OS starts up (step B2).

[0020] Therefore, by doing this, data changes on the main storage unit 13 will be correctly reflected in the secondary storage device 2, and the data in the secondary storage device 2 will be updated correctly, so that changes caused by power failure can be prevented. Data loss can be avoided. In this case, the backup power supply device 16 stores the necessary data in the main storage section 13.
Since power is only supplied to the computer main body 1 during the time it takes to save data from the data to the non-volatile memory 15, the capacity of the standby power supply device 16 can also be made small. For example, in the case of a workstation level computer system, the computer main body 1
The power consumption in the computer is about 800 [W], and the power consumption in the secondary storage device 2 is about 400 [W]. Therefore, considering a certain period of time, when backing up only the computer body 1, the power consumption in the secondary storage device 2 is about 400 [W]. 2, the capacity can be covered by the backup power supply device 16, which has about 2/3 of the capacity.Furthermore, in reality, access to the nonvolatile memory 15 is equivalent to access to the secondary storage device 2. Since the data saving can be performed at a higher speed, the time required for data saving can be shortened, and the capacity of the backup power supply device 16 can be further reduced. Furthermore, compared to the conventional system in which two sets of computer systems are provided, costs can be significantly reduced.

Next, FIG. 4 shows another embodiment of the present invention, in which a failure in the main power supply 3 is recovered and the system is started up. In this case as well, the initialization program is started first and refers to the variable S in the nonvolatile memory 15 (step C1). and,
No data has been saved to the nonvolatile memory 15, and S=
If it is 0, the determination is NO, and the OS is started according to the normal procedure (step C2). On the other hand, if the data has been saved in the non-volatile memory 15 and S=1, it is determined as YES and the data in the non-volatile memory 15 is written back to the main storage section 13 (
Step C3), then the variable S in the nonvolatile memory 15 is set to "0" (step C4), and then the OS is started (step C2).

[0022] In this way, when a power failure occurs, the data in the buffer for the secondary storage device of the main storage unit 13 is saved to the nonvolatile memory 15 and then transferred to the main storage unit 13 again. Since the data is returned to the buffer for the secondary storage device, faster processing can be obtained than writing back the data in the nonvolatile memory 15 to the secondary storage device 2, and the time required for system recovery can be shortened. .

Next, FIGS. 5 and 6 show another embodiment of the present invention. In this case, when a failure occurs in the main power supply unit 3 and the power supply to the computer body 1 is stopped, the 5 is executed, the backup power supply 16 starts operating as the main power supply 3 stops supplying power, supplies power to the computer main body 1 instead of the main power supply 3, and at the same time interrupts the CPU 11. multiply(
Step D1).

[0024] When the CPU 11 learns of the occurrence of a power failure through this interrupt, the CPU 11 stores the secondary storage device 2 of the main storage unit 13.
All data in the buffer for non-volatile memory 15
(Step D2). In this case, the nonvolatile memory 15 has the variable S initialized to "0" in advance, and is set to S=1 by data saving (step D3). Then, the computer main body 1 stops operating.

Next, when the failure of the main power supply device 3 is recovered and the system is started up, the flowchart shown in FIG. 6 is executed. In this case, the initialization program is started and refers to the variable S in the nonvolatile memory 15 (step E
1). Here, if the data has not been saved in the nonvolatile memory 15 and S=0, it is determined as NO, and the normal procedure is performed.
Start up S (step E2). On the other hand, if the data is saved in the nonvolatile memory 15 and S=1, YES
It is determined that the data in the non-volatile memory 15 is transferred to the main storage unit 1.
3 (step E3), then the variable S in the nonvolatile memory 15 is set to "0" (step E4), and then the OS is started (step E2).

In this way, when a power failure occurs, all the data in the buffer for the secondary storage device of the main storage section 13 is saved to the nonvolatile memory 15, and then the main storage section 13 is saved again. This not only guarantees the contents of the secondary storage buffer in the main memory unit 13, but also restores the context of the process (program) that was being executed at the time of the failure. Since the computer can be restored to the execution state), it is possible to restart the process from the middle of the process that the computer was executing at the time the failure occurred.

In each of the above-described embodiments, a case has been described in which a power failure occurs, but even if the power is intentionally turned off, the data in the main memory section 13 that must not be lost will still be lost, as described above. This will be guaranteed when the computer system starts up again. In this case, in conventional computers, the data in the buffer in the main memory is written back to the secondary storage device each time the computer stops its operation, but since there is no need for this, it can be done at high speed. The system can be stopped. In particular, in the case of the embodiments shown in Figures 5 and 6, the internal state of the computer at the time the power is turned off is completely saved, and even when the computer is turned on again, the state at the time the power was turned off is reproduced. Moreover, this process is performed without accessing the secondary storage device.
High-speed processing can be obtained, and as a result, a computer system that can stop and start the system very quickly can be realized.

[0028]

[Effects of the Invention] According to the calculation system of the present invention, when the power supply from the main power source to the computer body is cut off for some reason, power is supplied to the computer body from the standby power source instead. Since at least the data in the data buffer on the main memory that must not be lost can be saved to a nonvolatile storage means, data loss can be avoided even when the main power is cut off due to a power failure or the like. In addition, the backup power supply supplies power only during the time required to save the necessary data from the main memory to a non-volatile storage means, and since this power supply is only used for backing up the computer itself, the backup power supply can be used as a small It can be made into a large capacity. Furthermore, since there is no need to prepare two sets of computer systems, it is possible to significantly reduce costs.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a schematic configuration of an embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of the embodiment.

FIG. 3 is a flowchart for explaining the operation of the embodiment.

FIG. 4 is a flowchart for explaining the operation of another embodiment of the present invention.

FIG. 5 is a flowchart for explaining the operation of another embodiment of the present invention.

FIG. 6 is a flowchart for explaining the operation of another different embodiment.

[Explanation of symbols]

1... Computer body, 11... CPU, 12... System bus,
13...Main storage unit, 14...I/O interface, 15
...Nonvolatile memory, 16.. Backup power supply device, 2.. Secondary storage device, 3.. Main power supply device.

Claims (1)

[Claims] Claim 1: In a computer system in which processing results within the computer main body are temporarily stored in a data buffer on a main memory section using power supplied from a main power source, and then written to an external storage means, the power supply from the main power source is provided. a backup power supply that supplies power to the computer main body in place of the power supply when the power supply is cut off, and a non-volatile power supply that uses the power supplied from the backup power supply to save at least the data that must not be lost in the data buffer on the main memory. A computer system characterized by comprising a storage means.