JPH0113141B2 - - Google Patents

Description

[Detailed description of the invention]

[Field of Application of the Invention] The present invention relates to a centralized integrated computer system, and in particular to an information file transmission method suitable for efficiently recognizing an abnormality in a lower-level computer and importing an information file in the lower-level computer in a central computer. Regarding. [Background of the Invention] Conventional technology for information transmission will be explained using FIGS. 1 to 5. In a centralized integrated computer system consisting of central computers 1 and 2 and lower-level processing compatible computers 4 connected by an intercomputer linkage device 5, transmission line management processing 10 of the central computers 1 and 2 is performed.
The following describes the import of the lower-level processing-compatible computer information file 8, the management, and the response when the lower-level processing-compatible computer abnormality 9 occurs. The central computer system has a dual configuration of 1 and 2, and they each have a shared memory 3, within which a common information storage memory 3-1 is allocated for line management. In the figure, 6 indicates the message flow, 7 indicates the computer-shared memory interface, and 11 indicates the end. FIG. 2 describes the configuration and contents of the common information storage memory 3-1 for line management in the shared memory 3. The main storage memory 3-1 reserves one case for each lower-level processing computer 4, and stores the operating state of the lower-level processing computer 4, the storage start address and capacity of the information file 8. This storage memory 3-1 is read and written by an interface device 7 between the computer and the shared memory. Generally connected to a computer linkage device,
There are two types of information transmission between computers: a message communication method and a direct memory reference method in which the central computer unilaterally reads the memory of a computer that supports lower-level processing, but the former is particularly effective because of the various processing factors that occur within each computer. The latter method, which does not require hardware or software communication procedures, is particularly used for transmitting information files that have a relatively large capacity and require minimal transmission processing time. Now, when the transmission line management processing 10 takes in the computer information file 8 corresponding to lower-level processing, the latter is the case, but in any case, the common information for line management corresponding to the computer 4 is sent to the computer 4 via the shared memory interface device 7. Take out the contents of memory 3-1 and use calculator 4
After determining the operating status of the information file 8, it will be executed based on the address and capacity of the information file 8. In addition, when an abnormality 9 occurs in a computer corresponding to lower-level processing on the central computers 1 and 2 side, recognition is made by receiving a report from the computer 4 where the abnormality has occurred and controlling the line corresponding to the computer 4 via the shared memory interface device 7. The contents will be stored in the upper common information memory 3-1. In this case, there are two possible abnormalities in the lower-level processing computer 4: a software abnormality and a hardware abnormality including a power outage. However, if the computer 4 can operate software-wise, the central A report is made to the computer, and if it is inoperable, a report is made by a hardware interrupt. On the other hand, there is a method in which the central computers 1 and 2 actively check the operating status of the lower-level processing computers 4, such as periodically sending messages to the lower-level processing computers 4. If the system is in an abnormal state where it cannot respond, it will have to acknowledge the abnormality by response monitoring timeout, which will waste unnecessary time. However, in conventional centralized integrated computer systems, as a countermeasure against the disappearance of report information from lower levels, it has been essential for the central government to actively periodically diagnose the operating status of lower levels. This is because, in the case of a central computer multiplex system, when switching between the main system and the slave system, the paired computer temporarily performs an indirect process, and loses its ability as a computer during that time. This is because if a hardware abnormality report occurs from a lower level, that information will be lost. FIG. 3 shows the steps up to system switching when an abnormality 9 occurs in the central computer 1. Reference numeral 13 denotes an auxiliary storage device, which stores system programs for the main system and the slave system and various processing software. 1-1 and 2-1 are memories for operating each processing software including system software. 12 is an auxiliary storage multiple access device. Step 1 is the occurrence of an abnormality, and Step 2 is an initial period for switching the central computer 2 from the slave system to the main system processing. In step 3, the system is completely switched.
In other words, although it is a short time, the period of step 2 becomes a problem in the case of a report from a lower level, and in order to counter this, various information loss countermeasures are required. In any of the above cases, what is particularly important in a centralized integrated system is how to ensure that the central computer quickly recognizes the effectiveness of the transmission information file when an error occurs in the lower-level processing computer and transfers the error information file. The question is whether misuse can be avoided.
FIG. 4 is a diagram showing the above processing contents in a time chart format. In the figure, 11 is the central computer, 41 is the processing sequence of the lower-level processing computer, and △
is the lower-level information file 8 generated within the central computer.
This is the import timing. Also, each process |--|
The contents are distinguished by lowercase letters in brackets [ ]. In order to clarify the contents, the above-mentioned line management common information 3-1 and subordinate information file 8 are marked with □ in the time chart. In terms of processing, first, lower-order information file 8 import timing △
The contents of the common information memory are retrieved [a] for line management purposes, and in [b] the operation status of the lower-level computer 4 is determined based on the retrieved contents and the information file 8 is
The capacity address calculation is performed, and the target information file is fetched at [c]. [a], [b],
The process [c] is thereafter repeated every time the above-mentioned timing Δ occurs. In addition, the central computer 1 is
When the system starts up, it issues a message delivery wait command [e] ₀ to the line, and handles the reception of abnormality notification messages from lower-level processing computers. on the other hand,
When an error 8-9 occurs in a computer that supports lower-level processing, the error analysis process [h] operates, collects and analyzes the error information, and sends an error status message to the central computer in the message transmission process [k] to the central computer. Send. The central computer receives the abnormality notification from the lower level at [e] ₁ , and stores the abnormal situation in the common memory for line management at [d]. What should be noted here is that T1
It's time. T1 is the time required for the central computer to recognize an abnormality in the lower-level processing computer,
Shortening this T1 time as much as possible was a countermeasure to prevent misuse of the abnormality information file.
Technically, it is possible to shorten the time by raising the operating level of the processing logic related to abnormal communication to a higher level than the normal information file import processing logic, but it is impossible to completely reduce it to zero, and there is a limit. It was hot. In this way, abnormalities that can be recognized by the lower-level processing computer itself can be reported to the central computer through message communication or hardware interrupts, but abnormalities during information file transmission or runaway processing may occur. There was no response when the information file was destroyed, and the response was to shorten the T1 mentioned above. For example, there are already many methods for improving the reliability of transmitted information, but all of these methods are designed to deal with the occurrence of hardware defects on the transmission line; It is not a means to deal with such cases. As shown in FIG. 5, when the message information 6-1 is small in capacity, the thumb check data 6-2 is
If you add , it is possible to deal with information destruction, but the actual information file 8 has a relatively large capacity,
In addition, since the contents frequently change 8-1, it would be wasteful and meaningless to create a thumb check 8-2 correspondingly. This conventional technique has the following problems. (1) Since the line management information is stored in the shared memory, the information transmission function will either stop when the shared memory goes down, or the backup process to keep it from stopping becomes complicated. (2) In order to notify the central computer of an abnormality from a computer capable of lower-level processing using message communication, hardware interrupts, etc.
If the central computer multiplex system is being switched, the contents of the notification will disappear, making the message loss countermeasure processing at the central computer complicated. (3) Because the transmission of the information file and the abnormality recognition by the lower-level processing computer cannot be synchronized, there is a risk that the upper-level central computer may end up using the lower-level abnormal information file. [Object of the Invention] An object of the present invention is to provide a reliable and efficient information file transmission system. [Summary of the Invention] The main point of the present invention is a conventional information transmission method in a centralized computer system called control decentralization and information centralization. Management information related to the transmission information is not stored, but the computer operating status and management information related to the transmission information are stored in part of the transmission information in the memory of the lower-level computer, and when the upper-level computer takes in the lower-level transmission information, the operating status is stored. It is possible to diagnose the lower-level computer from the information, and at the same time, it eliminates the need for countermeasures against information loss during multi-system configuration control due to an abnormality in the upper-level common part, simplifying the transmission processing procedure. [Embodiments of the Invention] Details of the present invention will be explained using FIGS. 6 to 10. First, FIG. 6 will be explained.
Conventionally, the common information storage memory 3-1 for line management, which was managed in the shared memory 3 of the central computers 1 and 2, is divided for each lower-level process, and is used as a processing-compatible management information storage memory 4-1 for each lower-level process. Secure it in a compatible computer. The storage address of the same information 4-1 is the same in each lower computer, and the central computer has common management information storage memories 1-2, 2-2 that store the storage addresses and capacities in the lower computers. When the central computers 1 and 2 import the internal information file 8 for lower-level processing, they first import the common management information memory 1-2 or 2-2 stored in their own computers.
According to the contents, the contents of the processing management information memory 4-1 in the lower-level computer are imported, and then the target lower-level processing-compatible computer information file 8 is imported based on the contents.
After understanding the capacity, storage address, and operating status of the computer, the implementation will be carried out from now on. Next, an example of the contents of the processing correspondence management information and the contents of the processing on the central computer side will be explained using FIGS. 7 to 10. FIG. 7 shows the processing management information 4 in the lower processing processing computer 4.
The relationship and contents of 1 and related transmission information 8-3, 8-4, and 8-5 are described. Transmission information is classified into three types: information for high-speed transmission 8-3, information for low-speed transmission 8-4, and detailed information at the time of abnormality occurrence 8-5, and the storage area address and capacity of each is determined by processing management information 4-1. I remember it. In addition, the same management information 4
-1 holds information TOP1 and TOP2 that store the operating status of the computer 4 that stores the same information and the usability status of the transmission information 8-3 and 8-4.
TOP1 information, except for information lock, is set by the computer's system program or hardware when an abnormality is detected. The TOP2 information is a cyclic fixed-period counter that is updated at the lowest level of application, equivalent to a human pulse, and is used as abnormality detection information for responding to dynamic loops. In this embodiment, in order to improve the reliability of transmission information in addition to the TOP1 and TOP2 information described above, the transmission information file and information valid information called TOP3 are integrated. An example of TOP3 will be explained using FIG. The setting points for TOP3 are transmission information 8-3,
Let 8-4 be the beginning of each block divided into a certain processing unit 8-8. By setting specific information/CDEF, etc. in this information when the computer is started up, the first problem is information file content destruction due to software runaway 8-7, computer abnormality occurring during information transmission 8-6 8-9, etc. , runaway program, or TOP at the system program level
By rewriting the contents of 3, the central computer side can recognize an abnormality in the transmitted information at the time of information acquisition. Figure 9 shows a time chart of TOP1, TOP2, and TOP3 setting timings. [i] is an initial processing program that always runs when the computer system for lower-level processing starts, with 1 in TOP1, 0 in TOP2, and / in TOP3.
Initialize CDEF. [U] is a TOP2 update processing program whose periodic timer activation is set by [i], and the TOP2 update processing is executed at a low operation level by a general application program. If the application program [S] ₁ associated with the information file 8-n generates a program error during operation, the program error will be recognized by the hardware error detection logic [V] and the system program [S] 1 will be recognized by the hardware error detection logic [V]. O] Passed to processing. [O] analyzes the content of the abnormality and sets codes that match the content to TOP1 and TOP3.
On the other hand, if application program [S] ₂ related to information file 8-n becomes a dynamic loop while it is operating, [U] will no longer be able to operate, and TOP2 will no longer be updated.
Similar to the occurrence of a program error, the central computer can recognize an abnormality in the information file. FIG. 10 is a flowchart showing the information acquisition process on the central computer side. In process 80, the contents of the common information memory 3-1 for line management are retrieved, and the process of importing the management information storage memory 8 corresponding to lower-level processing is performed.
In step 2, the result of the import process is determined. However, in the process 82, it is determined whether or not the import has been normally completed in terms of hardware, and the operating state of the computer corresponding to the lower-order processing based on the imported contents TOP1 and TOP2 is determined in the processes 84 and 88. Thereafter, if there is no abnormality, high speed information 92, period determination 94 and low speed information 96
If an abnormality is detected, abnormality processing 86 and 90 are performed, and the processing ends. Incidentally, the validity of the information file is determined by the TOP3 incorporated in the information file 8, and is then executed by an application program in the central computer that actually uses the information file. [Effects of the Invention] According to the present invention, (1) By centrally managing information for line management including lower-level computers only at the information source, the information can be maintained even if the shared section between central computers is stopped. Does not affect transmission function. (2) By integrating the operating status of the lower-level computer and the transmission information, it is possible to simultaneously determine the operating status of the lower-level computer and the validity of the information file, which are closely related to the transmitted information file, achieving high-speed processing. and the reliability of the information itself improves. (3) Since message communication is not used to grasp the operating status of lower-level computers, a multiplexed configuration on the central computer side can be easily realized. The present invention is intended to realize the basic function of information transmission in a centralized integrated computer system consisting of a plurality of central computers and lower-level computers, and the procedure remains unchanged even when applied to a large-scale system. The combination of message communication and the present invention can be expected to have the effect of simplifying complex system configuration control processing including transmission lines that cannot be avoided in a centralized computer system. Specific examples of effects are shown in Table 1. Even if the central configuration is multiplexed like this configuration, the transmission of information files and lower-level operating status information is
Because the central side has the initiative, computers capable of lower-level processing can be given control without being aware of which way the main work central computer has changed, even if configuration control for main work switching occurs on the central computer side. It is possible to continue the work independently, and since the central computer does not need to notify the subordinates of its status, there is no need for extra message exchange, which reduces the load on the line and the configuration control processing of the entire system. Simplification becomes possible. Table 1 shows the effects of this method.

【table】

【table】 [Brief explanation of drawings]

1 to 5 are explanatory diagrams of the conventional technology, FIGS. 6 to 9 are explanatory diagrams of an embodiment of the present invention, FIG. The figure is a system diagram of the present invention. 1...Central computer, 2...Central computer, 3...Shared memory between central computers, 4...Lower processing compatible computer, 4
-1...Processing management information storage memory, 5...Inter-computer linkage device, 6...Message flow, message, 7...
Computer-shared memory interface, 8...Transmission information file.

Claims (1)

[Scope of Claims] 1. In an information transmission method of a centralized integrated system consisting of a central computer system and a lower-level computer system connected by a computer linkage device, a part of the information file existing in the lower-level computer system includes the lower-level computer system. The central computer system is provided with system operating state information, and the central computer system includes means for importing the subordinate information file, and means for checking transmission information for abnormalities based on the operating state information at the same time as the import. transmission method for information files. 2. In claim 1, the information file importing means is configured such that the central computer system takes initiative in the import timing and is based on a direct memory reference method, and the information file importing means is configured to take the initiative in the central computer system during multi-system configuration control of the central computer system. An information file transmission method characterized in that even if an abnormality occurs on a lower level side, the information file is imported after the configuration control is completed.