JPH035832A - System for processing fault of information processor - Google Patents

Abstract

PURPOSE:To continue a processing after fault restoration without increasing overhead in a normal processing by saving the content of a main storage part in an extension storage device, executing a memory saving processing and rewriting the content which is memory-saved in the main storage part after the initialization of hardware only when the fault occurs. CONSTITUTION:The content of the main storage part 12 is stored in the extension storage part 6 when the fault occurs in an information processor 10, and hardware of the information processor 10 is initialized in response to the memory saving. After the initialization, an initial program is loaded on the main storage part 12, and the content from the extension storage part 6 is written into the main storage part 12 by the initial program. The processing which is executed before the fault occurs is resumed based on rewritten information. Thus, the increase of overhead in the processing time is prevented, and the processing can be continued after the fault restores.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to a failure handling method for an information processing device, and particularly to a failure handling method for an information processing device that loads a program after initializing the hardware in the event of a failure and then restarts the device. Regarding.

Prior Art In this type of conventional failure handling method, information such as start and end information of processing is always stored in a magnetic disk device, which is an external storage device, while the information processing device is in operation, and when a failure occurs, the information is stored after rebooting. Processing is restarted based on this information.

Conventional failure handling methods like this require the above information to be stored in an external storage device at all times, and therefore always involve input/output operations in synchronization with processing, and are therefore not suitable for real-time processing. There is a drawback that there is no

OBJECTS OF THE INVENTION An object of the present invention is to enable processing to continue after failure recovery without increasing processing overhead.
An object of the present invention is to provide a failure handling method for an information processing device suitable for real-time processing.

Structure of the Invention The failure handling method according to the present invention comprises: an extended storage section that can be directly accessed from the main body of an information processing apparatus; a memory save means that stores the contents of the main storage section in the extended storage section when a failure occurs in the information processing apparatus; initializing means for initializing the hardware of the information processing device in response to memory save; loading means for loading an initial program into the main memory after this initialization; The present invention is characterized in that a memory recovery means for writing back to the main storage section is provided, and the processing that was being executed before the occurrence of the failure is restarted based on the written back information.

Another failure handling method according to the present invention is a failure handling method in an information processing system having a service processor that performs failure handling of an information processing device, wherein the information processing device includes a service processor that performs failure recovery when a failure occurs in the information processing device. Providing a failure notification means for notifying the service processor of processing,
In the service processor, memory save means stores the contents of the main memory of the information processing device upon receiving the failure notification, and initialization means initializes the hardware of the information processing device in response to the memory save. a load instructing means for instructing to load an initial program into the main memory after this initialization; a memory recovery means for writing back the contents from the memory save means to the main memory after loading the initial program; The information processing apparatus is characterized in that a restart means is provided for putting the information processing apparatus into an operating state after the rewriting, and the processing that was being executed before the occurrence of the failure is restarted based on the written back information. do.

Embodiments Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a functional block diagram and a schematic processing flow diagram of an embodiment of the present invention. When a failure occurs, memory save means 1, HW (hardware) initialization means 2, IPL
It is shown that the (initial program load) means 3, the memory recovery means 4, and the processing restart means 5 are activated in sequence.

Extended Memory
Reference numeral 6 denotes a semiconductor memory that can be directly accessed from the ↑n information processing device, and is provided in a place other than the main memory section, and is used to save the contents when a failure occurs in this main memory section.

The program file 7 stores an initial program in advance, and according to instructions from the IPL means 3, this initial program is read out and loaded into the main storage section within the information processing device.

FIG. 2 is a diagram showing the configuration of an information processing system to which the functional blocks of the embodiment shown in FIG.
0 is an instruction execution unit 11, a main memory unit 12, and an input/output control unit 15
It consists of ``In addition to the main storage section, it includes an execution status table 13 that stores the execution status of each process, and a program 14 that executes the process. Further, an extended storage device 6 is connected to the instruction execution unit 11, and the input/output control unit 1
A program file 7 is connected to 5.

FIG. 3 is a diagram showing the flow of interrupt processing performed by the program 14. FIG. 4 is a diagram showing the flow of system startup processing performed by the program 14.

Next, the normal operation of the information processing device 10 of this embodiment will be explained. The purpose of the program 14 is to perform online transaction processing, and receives processing requests from an external device (not shown) via the input/output control section 15. Data and a processing identification number are added to the processing request.

After the processing is completed, all processing is completed by transmitting the processing identification number and completion notification to the processing request source.

Here, until the end notification is sent, data can be obtained again by sending a process identification number and a retransmission request to the process request source, and the process can be re-executed.

The program 14 always stores the process request source identification, process identification number, and process progress status in the execution status table 13.

Next, failure processing after a failure occurs will be explained using FIGS. 3 and 4. A failure in the information processing device 10 is reported to software directly as a machine error interrupt or indirectly as an illegal interrupt.

The interrupt handling routine of program 14 begins with process 20.
It is determined whether the interrupt is due to a failure or not. If the interrupt is due to a failure, the contents of the execution state table 13 are stored in the extended storage unit 6 via the instruction execution unit 11 in step 21. Then, in process 22, the information processing device 1
An instruction to initiate a power-on reset of 0 is executed.

On the other hand, if the processing 20 determines that the interrupt is a normal interrupt, such as a 110 interrupt or a timer interrupt, the processing 20
3, normal interrupt processing is performed for each interrupt cause.

Now, information processing device 1 on which power-on reset has been activated
0 is an instruction execution unit 11, a main storage unit 12, and a human output control unit 1.
5, automatically starts the input/output control section 15, reads the program 14 from the program file 7, stores it in the main storage section 12, and starts the first instruction of the program 14.

The activated program 14 reads the contents of the extended storage device 6 into the work area of the main storage unit 12 in process 30, and determines whether valid data is stored in process 31. If stored, the data read out to the work area is transferred to the execution state table 13 in process 32, invalid data is written to the extended storage device 6 in process 33, and stored in the execution status table in process 34. The suspended process is resumed based on the information obtained.

On the other hand, if it is determined in process 31 that the execution status table is invalid, the process proceeds to process 34, and the process is started using the execution state table in the initial state.

Next, another embodiment of the present invention will be described. In this embodiment, the service processor (
This is an example of a case where this is performed using SVP).

FIG. 5 is a functional block diagram thereof and a schematic processing flow diagram. The information processing device performs SV in response to the occurrence of a failure.
It includes a fault notification means 51 that notifies P of this fact, and a processing restart means 57 that restarts the processing upon receiving a recovery processing restart activation command from the SVP.

In the SVP, upon receiving the notification from the failure notification means 51, the memory save means 52, the HW initialization means 53, and the IP
This shows that the L means 54, the memory recovery means 55, and the restart means 56 are activated in sequence.

The save file 8 is for saving the contents of the main memory of the information processing device when a failure occurs.

The program file 59 stores an initial program in advance, and according to instructions from the IPL means 54, this initial program is read out and loaded into the main storage section within the information processing device.

FIG. 6 is a diagram showing the configuration of an information processing system to which the functional blocks of the embodiment shown in FIG.
0 is an instruction execution unit 61, a main memory unit 62, and an input/output control unit 65
It consists of The main storage unit 62 includes an execution status table 63 that stores the execution status of each process, and a program 64 that executes the process. Further, a peripheral device group 66 and a program file 59 are connected to the input/output control section 65.

S V P 67 C S V P instruction execution unit 68 and s
It consists of a vp memory 69 and an SVP input/output control unit 71, and the SvP memory 69 stores a fault handling program. The information processing device 60 and the 5VP 67 are connected through a diagnostic interface 72, and the save file 58 is connected to the SvP input/output control unit 71.

The normal operation of the information processing device 60 in this embodiment is the same as the normal operation of the information processing device 10 in the embodiment shown in FIG. 2, and the explanation thereof will be omitted.

FIG. 7 is a diagram showing the flow of interrupt processing performed by the program 64, and failure processing after a failure occurs will be explained with reference to FIGS. 5 to 7.

A failure of the information processing device 60 is reported to software directly as a machine error interrupt or indirectly as an illegal interrupt. In the interrupt processing routine of the program 64, first, in process 73, it is determined whether or not the interrupt is due to a failure.If the interrupt is due to a failure, then in process 74, the diagnostic interface
The 5VP 67 is notified via the 5VP 67, and in process 75 it is brought into a stopped state.

On the other hand, when it is determined in process 73 that the interrupt is a normal interrupt, such as a 110 interrupt or a timer interrupt, normal interrupt processing for each cause of the interrupt is performed in process 76.

Now, the information processing device 60 via the diagnostic interface 72
The 5VP 67 receives the failure notification and starts the failure handling program 70. Started failure handling program 70
The execution status table 63 stored in the main storage section 62 is saved by the memory save means 52 to the diagnostic interface 72.
, the SvP input/output control unit 71 is activated, and the data is stored in the save file 58. Then, the information processing device 60 is configured as the diagnostic interface 7 by the HW initialization means 53.
2, and the IPL means 54 issues an IPL request to the human output control section 65.

Here, the human output control unit 65 that has received the IPL request reads the program 64 and the initial values of the execution state table 63 from the program file 59, stores them in the main storage unit 62,
Reports completion of IPL to 5VP67.

Upon receiving the completion report, the failure processing program 70 uses the memory recovery means 55 to read out the information in the execution state table 63 stored in the file 58 and writes it back to the main storage section 62 . Thereafter, the processing restart means 56 instructs the information processing device 60 to start up via the diagnostic interface 72,
Terminate failure handling.

Now, the activated information processing device 60 executes processing from the first command of the program 64. That is, the processing resuming means 57 inspects the execution state table 63, and if it is in the initial state, it is determined that there is no processing to be restored, and if it is not in the initial state, the suspended processing is resumed based on that information. .

In each of the embodiments described above, a fault handling method for online transaction processing has been described, but it is clear that the present invention can be applied to all information processing such as batch processing.

Effects of the Invention As described above, according to the present invention, only when a failure occurs, the contents of the main memory at that time are transferred to the expanded storage device or SV.
Since the memory save processing is performed by evacuating the data to P, and the memory-saved content is written back to the main memory after HW initialization, the overhead in normal processing does not increase, and processing can be continued after failure recovery. There is an effect.

[Brief explanation of drawings]

Figure 1 is a functional processing flow diagram of an embodiment of the present invention, Figure 2 is a system configuration diagram to which the embodiment of the present invention is applied, Figure 3 is a flow diagram of interrupt processing, and Figure 4 is a flow diagram of system processing. , FIG. 5 is a functional processing flow diagram of another embodiment of the present invention,
FIG. 6 is a system configuration diagram to which the embodiment of FIG. 5 is applied;
FIG. 7 is a flowchart of interrupt processing in the embodiment of FIG. Explanation of symbols of main parts 1.52...Memory saving means 2.53...
... HW initialization means 3.54 ... IPL means 4.55 ... Memory recovery means 5.57 ...
. . . Processing restart means 6 . . . Extended storage device 7.59 . . . Program file 8 .
・Memory save file Figure 1 6...Restart method

Claims (2)

[Claims] (1) An extended storage section that can be directly accessed from the information processing device main body, a memory save means that stores the contents of the main storage section in the extended storage section when a failure occurs in the information processing device, and information processing in response to the memory save. initializing means for initializing the hardware of the device; loading means for loading an initial program into the main memory after the initialization; and writing the content from the expanded memory back to the main memory using the initial program. 1. A failure handling method for an information processing apparatus, characterized in that a memory recovery means is provided, and processing that was being executed before the occurrence of the failure is restarted based on the written back information. (2) A failure handling method in an information processing system having a service processor that performs failure handling of an information processing device, wherein the information processing device notifies the service processor of failure recovery processing when a failure occurs in the information processing device. a failure notification means for storing the contents of the main memory of the information processing device in response to the failure notification; an initialization means for initializing the initial program; a load instruction means for instructing to load an initial program into the main memory after the initialization; and a load instruction means for instructing the initial program to be loaded into the main memory, and after loading the initial program, the contents are written back from the memory save means to the main memory. A memory recovery means and a restart means for bringing the information processing device into an operating state after this write-back are provided, and the processing that was being executed before the occurrence of the failure is resumed based on the written-back information. A fault handling method for an information processing device, characterized in that: