JPH08314785A - Data file equalization processing method - Google Patents

Abstract

PURPOSE: To surely equalize data files in a short time without damaging the performance of a dual system operation of current system and standby system at the time of starting a standby-system computer. CONSTITUTION: For example, when a computer 2 is used as the current-system computer and a computer 1 is used as the standby-system computer in a dual computer system, computers 1 and 2 of both systems asynchronously count up the update frequency counters in the unit of records and store the current- system computer discrimination information for update after a data file 221 of the current-system computer 2 is updated, and update frequency counters of computers 1 and 2 at the time of the start of the standby system or only records different by current-system computer discrimination information for update are copied from the current-system computer 2 to the standby-system computer 1. Thus, data files are surely equalized in a short time without damaging the performance of dual system operation of the current system and the standby system at the time of starting the standby-system computer 1.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data file equivalent processing method in an operating computer and a standby computer of a multiplexed computer system.

[0002]

2. Description of the Related Art Conventionally, an information processing system and a control processing system using a computer system are generally provided with an operating computer and a standby computer as a dual system configuration in order to improve their reliability. You can see the one that operates the computer and performs the duplex operation. In such a system, every time the data file is updated on the operating computer,
By writing the same data to the data file of the standby computer, the data files of both systems are made equivalent. Also,
When starting up to the standby system after one computer has stopped due to a failure or maintenance, copy the file contents of the active computer to the data file of the new standby computer to match the data files of the computers of both systems. However, if the amount of target data is large, there is a problem in that copying takes time and, as a result, the standby startup time becomes long, and it is necessary to shorten the copy time.

FIG. 26 shows, for example, Japanese Patent Laid-Open No. 3-137742.
FIG. 27 is a hardware configuration diagram showing a two-series computer system that employs a conventional data file equivalence processing method for the purpose of shortening the copy time, and FIG. 27 is an example of a conventional data file equivalence processing method. It is a block diagram showing. In FIG. 26 and FIG. 27, reference numerals 1 and 2 denote computers constituting a double system, and 5 denotes a transmission line connecting the computers 1 and 2. 10 is the CPU of the computer 1, 11 is the main memory of the computer 1, 12 is the disk of the computer 1, 121 and 122
Is a data file in the disk 12, and 13 is a transmission interface unit that connects the computer 1 and the transmission line 5. 20
Is the CPU of the computer 2, 21 is the main memory of the computer 2, 22
Is a disk of the computer 2, 221 and 222 are disks 22
A data file 23 is a transmission interface unit that connects the computer 2 and the transmission line 5. 1211 is a record in the data file 121, 1221 is a record in the data file 122, 111 is a record unit of the data file indicating whether the data file of the computer 1 is updated when the computer 1 is the active system and the computer 2 is stopped. It is a management unit that is stored in the main memory 11. 221
1 is a record in the data file 221, 2221 is a record in the data file 222, 211 is a record unit of the data file indicating whether or not the data file of the computer 2 is updated when the computer 2 is the active system and the computer 1 is stopped. It is a management unit that is stored in the main memory 21. The record 1211 and the record 2211 correspond to each other, and the record 1221 and the record 2221 correspond to each other. When one of the computer 1 and the computer 2 operates as an active system and the other operates as a standby system, the contents of the corresponding records must always be the same.

Next, the operation will be described. When the computer 2 is the active system, the active computer 2 performs the data change process (active system) according to the flowchart shown in FIG. That is, when the operational computer 2 receives the data update request, the data file 2 of the computer 2 is calculated in step ST501.
21 or the data file 222, the update data is written, and it moves to step ST502. In step ST502, it is checked whether or not the other system computer is the standby system. If it is the standby system, the process proceeds to step ST503, and if it is not the standby system, the process proceeds to step ST504. In step ST503, the update data is transferred to the computer 1 and the data change process is completed. In step ST504, the fact that the record has been updated is stored in the management unit 211 of the operating system computer 2, and the data change processing ends. According to the flowchart shown in FIG. 28, for example, if the active computer 2 updates the record 2211 and the record 2221 when the computer 1 is not in the standby mode, the record 221 is stored in the computer management unit 211.
1 and the record 2221 are stored as updated. Further, for example, if the computer 2 is the active system and the computer 1 is the standby system, even if the record 2211 or the record 2221 is updated by the active computer 2, it is not stored in the management unit 211 of the active computer 2.

If the computer 1 is a standby system, the standby system computer 1 performs a data change process (standby system) according to the flow chart shown in FIG. That is, as a result of step ST503 of FIG. 28, when the data change process (standby system) is started when the update target data is received from the active system computer 2, the data is changed to the data file 121 or the data file in step ST601 of FIG. Data is written in the corresponding record 122, and the data change process (standby system) is completed. 28 and 29, for example, assuming that the computer 2 is the active system and the computer 1 is the standby system, if the record 2211 and the record 2221 are updated by the active computer 2, the standby computer has the same contents as the record 2211. The record 1211 of No. 1 is updated, and the record 1221 of the standby computer 1 is updated with the same contents as the record 2221.

Next, when the stopped computer 1 is booted to the standby system, the active computer 2 performs a data copy process (active system) according to the flowchart shown in FIG.
That is, in step ST701, the first record of the data file is set as the check target record, and the process proceeds to step ST702. In step ST702, the management unit 211 of the computer 2 is referenced to determine whether the check target record has been updated when the computer 1 is not in the standby system, and when the data update is stored, step ST7
03, and when the data update is not stored, it moves to step ST705. In step ST703, the contents of the check target record are read from the data file 221 or the data file 222, and step ST70
Go to 4. In step ST704, step ST70
The data read in 3 is transferred to the computer 1 being started up, and the process proceeds to step ST705. In step ST705, it is determined whether or not an unprocessed record remains, and if there is any record, the process proceeds to step ST706, and if there is no record, the data copy process (active system) ends. Step ST
In 706, the record to be checked is the next record,
Move to step ST702.

When the stopped computer 1 is brought up to the standby system, the computer 1 of the new standby system performs the data copy process (new standby system) according to the flowchart shown in FIG. That is, as a result of step ST704 in FIG.
When the data copy process (new standby system) is started when the data to be copied is received from the operating system computer 2, the data is copied to the corresponding record of the data file 121 or the data file 122 in step ST801 of FIG. , Data copy processing (new standby system) ends. 29 and 30, for example, if the management unit 211 stores that the record 2211 and the record 2221 are updated, the record 2211 and the record 2221 are stored.
Data is transferred to the computer 1 and copied to the record 1211 and the record 1221, respectively.

[0008]

According to the conventional data file equivalence processing method, for example, the operating system computer 2 operates as shown in FIG.
8 in step ST501, update record 2211,
In step ST502, the computer 1 is determined to be in the standby mode, the update data is transferred to the computer 1 in step ST503, and then the standby system computer 1 executes the step ST601 in FIG.
If the computer 1 starts transitioning to a state other than the standby system before the execution of
11 is updated and the record 1211 of the computer 1 is not updated, the record 2 is stored in the management unit 211.
A phenomenon occurs in which the fact that 211 has been updated in the independent system is not stored. After that, when the computer 1 is started up in the standby system, in FIG. 30, it is determined that the data update of the record 2211 is not stored in step ST702, and steps ST703 and ST704 are performed.
Record 2211 of the active computer 2 and the corresponding record 1211 of the standby computer 1 in order to skip the above process.
There was a problem that the contents of did not match.

In order to prevent the above problems, after the process of step ST601 shown in FIG. 29, a process of transmitting the fact that the writing of the record to the standby system is successful to the active system is added, and the step shown in FIG. After ST502, it may be possible to wait for the reception of the write result from the other system and, if the write in the other system is not successful, change to skip step ST504. Every time it is changed, the processing performance deteriorates because it synchronizes with the data writing in the standby system.
There is a problem.

Further, in the conventional method, since the management unit for storing the presence / absence of data update is in the main memory, one computer is in the operating system, the other computer is in the standby system, and the standby computer is stopped. If the active computer stops while the status is in the state of, the contents of the management section will be cleared, so when one computer returns to the active system and the other computer returns to the standby system, the data file is updated from the active computer. There is a problem that you have to copy all to the standby computer.

The present invention has been made in order to solve the above problems, and starts up a standby computer without impairing the performance of the dual system operation or the multiple system operation in the active system-standby system. It is an object of the present invention to obtain a data file equalization processing method that can surely equalize data files in a short time.

[0012]

According to the data file equivalence processing method of the invention of claim 1, in the dual computer system, the equivalence target data files 121 and 221 of the respective computers 1 and 2 are recorded in record units. A management unit 1 that stores an update counter and an operating system computer identification information at the time of update
20 and 220 are included in the disks 12 and 22, and when the equivalence target data file is updated in the active computer or the standby computer, the update counter and the active computer identification information at the time of update are updated, When a computer that has stopped due to a failure or maintenance starts up and transitions to the standby system, copy only the records that have different update counters on both systems or the active computer identification information when updating from the active computer to the new standby computer. It was done like this.

A data file equivalence processing method according to a second aspect of the present invention is an equalization target data file 12 of each computer 1, 2 and 3 in a multi-system computer system.
The management unit 12 that stores the update counter and the operating-system computer identification information at the time of update in units of 1, 221, and 321 records.
0, 220, 320 in the disks 12, 22, 32, and after updating the equivalence target data file in the active computer or the standby computer, the update count counter corresponding to the record and the active computer identification information at the time of update When a computer stopped due to a failure or maintenance starts up and transitions to the standby system, only the records with different update counters of both systems or the operating system computer identification information at the time of updating are updated from the active computer to the new standby system. It is designed to be copied to a computer.

[0014]

According to the data file equivalent processing method of the first aspect, in the operating system computer and the standby system computer of the dual system computer system, for example, if the computer 2 is the operating system and the computer 1 is the standby system, the operating system is After updating the data file 221 of the computer 2, the computers 1 and 2 of both systems asynchronously count up the update counter in record units and store the operating system computer identification information at the time of update, and the computer at the time of starting the standby system. Only the records in which the update count counter of 1 and the active computer 2 or the active computer identification information at the time of update differ are copied from the active computer 2 to the new standby computer 1. As a result, the data files are reliably equalized in a short time when the standby system computer 1 is started up, without impairing the performance of the double operation in the active system-the standby system.

In the data file equivalent processing method according to the second aspect of the present invention, in the operating system computer and the standby system computer of the multi-system computer system, for example, if computers 1 and 2 are the operating system and computer 3 is the standby system, After updating the data file 221 of the active computer 2, the computers 2 and 3 of both systems asynchronously count up the update counter in record units, store the active computer identification information at the time of update, and start up the standby system. Only the records in which the update count counters of the computer 3 and the active computer 2 at the time or the active computer identification information at the time of update are different are copied from the active computer 2 to the new standby computer 3. As a result, the data files are reliably equalized in a short time when the standby system computer 3 is started up without impairing the performance of the multiple system operation between the active system and the standby system.

[0016]

【Example】

Example 1. Embodiment 1 of the present invention will be described below with reference to the drawings. 1 is a hardware configuration diagram showing a dual computer system adopting a data file equivalent processing method according to a first embodiment of the present invention, and FIG. 2 is a block diagram showing a data file equivalent processing method according to the first embodiment. is there. Figure 1
In FIG. 2, reference numerals 1 and 2 denote computers constituting a dual system, and 5 denotes a transmission line on a network connecting the computers 1 and 2. Reference numeral 10 is a CPU of the computer 1, 11 is a main memory of the computer 1, 12 is a disk of the computer 1, 121 and 122 are data files in the disk 12, and 13 is a transmission interface unit that connects the computer 1 and the transmission path 5. Reference numeral 20 is a CPU of the computer 2, 21 is a main memory of the computer 2, 22 is a disk of the computer 2, 221 and 222 are data files in the disk 22, and 23 is a transmission interface unit that connects the computer 2 and the transmission path 5. 1211 is a record of the data file 121, 1221 is the data file 12
2 records, 2211 is a record of the data file 221, and 2221 is a record of the data file 222. Record 1211, record 2211, record 1
221 and record 2221 correspond to each other, and when one of computer 1 and computer 2 operates as an active system and the other operates as a standby system, the contents of the corresponding records must always be the same. A management 120 has an update counter that is incremented in record units each time the data file of the computer 1 is updated when the computer 1 is the active system or the standby system, and the active computer name (active computer identification information) at the time of update. And is present inside the disk 12. 220
Is a management unit having an update counter that is incremented in record units each time the data file of the computer 2 is updated when the computer 2 is the active system or the standby system, and the operating system computer name at the time of update, and is present inside the disk 22. To do.

FIG. 3 is a diagram showing the structure of the management unit 120. In FIG. 3, 12110 is a data file 121.
Update counter 1211 of the record 1211, 12111 is the operating computer name when the record 1211 was updated, 1
Reference numeral 2210 is the update count counter of the record 1221 of the data file 122, and reference numeral 12211 is the operating computer name when the record 1221 was updated. FIG. 4 shows the management unit 22.
4 is a diagram showing the structure of 0. In FIG. 4, 22110 is the update counter of the record 2211 of the data file 221, 22111 is the name of the operating computer when the record 2211 was updated, and 22210 is the data file 222.
The update number counter of the record 2221 and 22211 are the operating system computer names when the record 2221 was updated. The update counter shall be capable of expressing a sufficiently large value, such as a 4-byte unsigned integer, shall be cleared to 0 at the start of the first operation of the system, and shall be reset to 0 when it is counted up at the maximum expressible value. And

Next, the operation will be described. The active computer performs data change processing (active) according to the flowchart shown in FIG. That is, when the working computer receives the data change request, it writes the update data to the data file of its own system in step ST101, and then in step ST1.
Move to 02. In step ST102, the update number counter corresponding to the update target record in the management unit existing in the disk of the active computer is counted up, and the process proceeds to step ST103. In step ST103, the self-system computer name is written in the update-time active system computer name corresponding to the update target record in the management section of the active system computer, and the process proceeds to step ST104. In step ST104, it is checked whether or not there is a standby system computer other than the own system. If there is a computer, the process proceeds to step ST105, and if not, the data change process (active system) ends. In step ST105, the update data is transferred to the standby computer and the data change process (active system) ends.

The standby system computer performs data change processing (standby system) according to the flow chart shown in FIG. That is,
As a result of step ST105 in FIG. 5, when the update target data is received from the active computer, the data change process (standby system)
When is activated, in step ST201 of FIG. 6, the data is written in the corresponding record of the data file, and the process proceeds to step ST202. Step ST202
Then, the update number counter corresponding to the update target record in the management unit existing in the disk of the standby computer is counted up, and the process proceeds to step ST203. In step ST203, the name of the other system computer is written to the update-time operating system computer name corresponding to the update target record in the management unit of the standby system computer, and the data change process (standby system) ends.

Next, when the stopped computer is booted to the standby system, the active computer performs the data copy process (active system) according to the flowchart shown in FIG. That is, in step ST301, the first record of the data file is set as the check target record, and step ST
Move to 302. In step ST302, the update count counter of the check target record of the management unit of the own system and the update count counter of the check target record of the management unit of the standby transition type computer are compared. If they are equal, the process proceeds to step ST303, and if they are not equal, step ST304. Move to. In step ST303, the update-time operating system computer name of the check target record of the management unit of the own system and the update-time operating system computer name of the check target record of the management unit of the standby transition system computer are compared. If they are equal, the process proceeds to step ST306. If not equal, the process proceeds to step ST304. Step ST
In 304, the contents of the check target record are read from the data file, and the process proceeds to step ST305. In step ST305, the data read in step ST304 is transferred to the computer in the transition to the standby state, and step ST30
Go to 6. In step ST306, it is determined whether or not there is an unprocessed record remaining, and if there is any unprocessed record, step S306 is performed.
The process moves to T307, and if there is no remaining data, the data copy processing (active system) ends. In step ST307, the check target record is set as the next record, and step ST302
Return to.

The computer of the new standby system, when the computer 1 which has been stopped is started up to the standby system, performs the data copy process (new standby system) according to the flowchart shown in FIG.
That is, as a result of step ST305 in FIG. 7, when the data copy processing (new standby system) is started when the copy target data is received from the active computer, step ST4 in FIG.
In 01, the data is written in the corresponding record of the data file, and the process proceeds to step ST402. In step ST402, the update counter and the update-time active computer name of the management unit of the active computer corresponding to the processed record are copied to the corresponding area of the management unit of the own system computer, and the data copy process (new standby system) is performed. ) Is finished.

For example, it is assumed that the computer 1 is operating in a standby system and the computer 2 is operating in an active system. It is assumed that the data file and the management unit at this time have the contents shown in FIG. That is, the content of the record 1211 of the data file 121 of the computer 1 is A, and the record 122 of the data file 122 of the computer 1 is
1 is X, the value of the update count counter 12110 of the record 1211 of the data file 121 of the computer 1 is a,
When the update-time operating system computer name 12111 is “Computer 1”, the update-counter 12210 of the record 1221 is “b”, and the update-time operating system computer name 12211 is “Computer 2”, the operating system + standby system is operated as a dual system. Therefore, since the contents of the data files of the computer 1 and the computer 2 and the corresponding areas of the management unit are the same, the contents of the record 2211 of the data file 221 of the computer 2 are A and the contents of the record 2221 of the data file 222 of the computer 2. Is X, the value of the update number counter 22110 of the record 2211 of the data file 221 of the computer 2 is a, the update-time operating system computer name 22111 is “computer 1”, the update number counter 22210 of the record 2221 is b, the update-time operating system computer The name 2221 is "Computer 2". When the record 2211 of the data file 221 of the active computer 2 is updated to B in this state, the record 121 of the data file 121 of the standby computer 1 is obtained as a result of the processing of FIGS. 5 and 6.
1 is updated with the same data B, the update number counter 12110 of the standby computer 1 and the update number counter 22110 of the active computer 2 become a + 1, and the update-time active computer name 12111 of the standby computer 1 and the active computer The update-time operating system computer name 22111 of 2 is “computer 2” (FIG. 10). Next, the record 2221 of the data file 222 of the active computer 2 was updated with the data Y, but if the standby computer 1 is stopped due to a failure or the like immediately before step ST201 of FIG. The record 1221 of the file 122 is not updated, and the update counter 22210 of the active computer 2 becomes b + 1, but the update counter 12210 of the stop transition computer 1 remains b and the update operation of the active computer 2 is performed. The system computer name 22211 is overwritten by "computer 2", and the update-time operating system computer name 1 of the stop transition system computer 1
2211 remains “Computer 2” (FIG. 11). Next, even if the active computer 2 is stopped, the data files of both computers, the update counter, and the update active computer name do not change (FIG. 12). Next, even if the computer 2 is started up as an active system, the contents of the data file and the management unit will not change (FIG. 13). Next, when the computer 1 is started up in the standby system, FIG.
8 operates, the update count counter 12110 of the new standby system computer 1 and the update count counter 22110 of the active system computer 2 are equal, and the update-time active system computer name 12111 and the active system computer 2 of the new standby system computer 1 Since the operating system name 22111 at the time of updating is the same, the new standby system computer 1
The contents of the record 1211 of the operating system computer 2 remain unchanged.
The update number counter 22210 of the new standby system computer 1 and the update number counter 12210 of the new standby system computer 1 are not equal to each other. Therefore, the content Y of the record 2221 of the active system computer 2 is the new standby system computer 1
Of the update number counter 22210 of the active computer 2 to the update number counter 12210 of the new standby computer 1, and the contents of the update active computer name 22211 of the active computer 2 to the new standby The updated system computer name 12211 of the system computer 1 is copied, and the contents of the data files of both systems and the management unit become the same (FIG. 14).

Further, even if the standby computer 1 is stopped in the state of FIG. 9 due to a failure or the like, the data files of both computers, the update counter, and the update operating computer name do not change (FIG. 15). When the record 2211 of the data file 221 of the active computer 2 is updated to B in this state, the update count counter 22110 of the active computer 2 becomes a + 1 as a result of the operation of FIG. 5, and the update operation of the active computer 2 is performed. The system computer name 22111 is "computer 2",
Data file of stop computer 1, update counter,
The operating computer name at the time of update does not change (Fig. 16). Next, even if the active computer 2 is stopped due to a failure or the like, the data files of both computers, the update counter, and the update active computer name do not change (FIG. 17). Next, even when the computer 1 is booted to the operation mode, the data files of both computers 1 and 2, the update count counter, and the update-time active computer name do not change (FIG. 18). When the record 1211 of the data file 121 of the active computer 1 is updated to C in this state, the update count counter 12110 of the active computer 1 becomes a + 1 as a result of the operation of FIG. 5, and the update operation of the active computer 1 is performed. The system computer name 12111 is “computer 1”, but the data file of the stopped system computer 2, the update count counter, and the update-time operating system computer name do not change (FIG. 19). Next, when computer 2 is started up in the standby system,
7 and 8 operate, the update count counter 22110 of the new standby computer 2 and the update count counter 12110 of the active computer 1 are the same, but the update standby computer name 22111 of the new standby computer 2 and the operation Since the active computer names 12111 at the time of updating the active computer 1 are not equal, the content B of the record 1211 of the active computer 1 is copied to the record 2211 of the new standby computer 2 and the update counter 12110 of the active computer 1 Is copied to the update counter 22110 of the new standby system computer 2, the contents of the active system computer name 12111 at the time of update of the active computer 1 are copied to the update active computer name 22111 of the new standby system computer 2, and the active computer 1 update counter 122
10 and the update count counter 22210 of the new standby computer 2
Are the same, and when updating the active computer 1, the active computer name 1
Since the update standby system computer name 22211 of 2211 and the new standby system computer 2 are the same, the contents of the record 2221 of the new standby system computer 2 do not change, and as a result, the data files of both systems are the same as the contents of the management unit ( (Fig. 20).

As described above, according to the first embodiment, after the record of the data file of the active computer 2 is updated, the standby computer 1 is stopped due to a failure or the like until the corresponding record of the standby computer 1 is updated. Even when the standby system computer 1 starts up in the standby system again, the corresponding records in the data files of the computers 1 and 2 of both systems are recognized as inconsistent and are copied from the active system computer 2 to the new standby system computer 1. In addition, since the management unit is on the disk, the contents of the management unit are not erased even if the computers 1 and 2 of both systems are stopped, and the name of the operating system computer at the time of updating is stored. Even after updating with a different value when computer 2 is the active system and computer 1 is the standby system, and then this time is updated with another value when computer 1 is the active system and computer 2 is the standby system, it was run again with the active system + standby system At this time, the records in the data files of the computers 1 and 2 of both systems are recognized as inconsistent and are copied from the active computer 1 to the new standby computer 2. As a result, no matter how the operating mode of the computers 1 and 2 of both systems transits, the data file record in which the contents of the computers 1 and 2 of both systems do not match when the standby system is started up can be correctly recognized, Since only the active computer 1 is copied to the new standby computer 2, the standby startup time is shortened.
In addition, when the active system-standby system is operating in a dual system, it is not necessary to synchronize with the data update in the standby system computer in the data change process of the operating system computer. Does not spoil.

In the first embodiment, for the sake of convenience, the computer 1 is used as the standby system and the computer 2 is used as the active system. However, as is clear from the above description, the computer 1 may be used as the active system and the computer 2 may be used as the standby system. Of course, also in this case, the same effect as that of the first embodiment can be obtained.

Example 2. The second embodiment of the present invention will be described below with reference to the drawings. 21 is a hardware configuration diagram showing a multi-system computer system adopting the data file equivalent processing method according to the second embodiment of the present invention, and FIG. 22 is a block diagram showing the data file equivalent processing method according to the third embodiment. In FIG. 21 and FIG. 22, reference numerals 1, 2 and 3 are computers constituting a triple system, and 5 is a transmission line on a network connecting the computers 1, 2 and 3. Reference numeral 10 is a CPU of the computer 1, 11 is a main memory of the computer 1, 12 is a disk of the computer 1, 121 and 122 are data files in the disk 12, and 13 is a transmission interface unit that connects the computer 1 and the transmission path 5. 20 is the CP of computer 2
U and 21 are main memories of the computer 2, 22 is a disk of the computer 2, 221 and 222 are data files in the disk 22, and 23 is a transmission interface unit which connects the computer 2 and the transmission path 3. Reference numeral 30 is a CPU of the computer 3, 31 is a main memory of the computer 3, 32 is a disk of the computer 2, 321 and 322 are data files in the disk 32, and 33 is a transmission interface unit that connects the computer 3 and the transmission path 5. 1211 is a record of the data file 121, 12
21 is a record of the data file 122, 2211 is a record of the data file 221, 2221 is a record of the data file 222, 3211 is a data file 32.
1 record, 3221 is a record of the data file 322. Record 1211, record 2211 and record 3211, record 1221 and record 2221
And record 3221 correspond to each other, and when any one of Calculator 1, Calculator 2 and Calculator 3 operates as an active system and the remaining 1 or 2 operates as a standby system, The contents of the corresponding records must always be equal. A management unit 120 has an update counter that is incremented in record units each time the data file of the computer 1 is updated when the computer 1 is the active system or the standby system, and the active computer name (active computer identification information) at the time of update. And exists inside the disk 12. Two
Reference numeral 20 denotes a management unit having an update counter that is incremented in record units each time the computer 2 updates the data file of the computer 2 when the computer 2 is the active system or the standby system, and the operating computer name at the time of update. Exists. Reference numeral 320 denotes a management unit having an update counter that is incremented in record units each time the data file of the computer 3 is updated when the computer 3 is an active system or a standby system, and an operating computer name at the time of update, and is present inside the disk 32. To do.

FIG. 23 is a diagram showing the structure of the management unit 120. In FIG. 23, reference numeral 12110 denotes the data file 12.
Update counter of record 1211 of 1; 12111
Is the name of the operating computer when the record 1211 was updated,
12210 is a record 1221 of the data file 122
The update number counter of 1211 is the operating system computer name when the record 1221 is updated. FIG. 24 is a diagram showing the structure of the management unit 220. In FIG.
10 is the update count counter of the record 2211 of the data file 221, 22111 is the name of the operating system computer when the record 2211 was updated, 22210 is the update count counter of the record 2221 of the data file 222, 222
Reference numeral 11 is the operating system computer name when the record 2221 is updated. FIG. 25 is a diagram showing the structure of the management unit 320. In FIG. 25, reference numeral 32110 denotes the data file 32.
Update counter of record 3211 of 1; 32111
Is the name of the operating system computer when the record 3211 was updated,
32210 is a record 3221 of the data file 322
The update number counter 32211 is the operating system computer name when the record 3221 is updated. The update counter is capable of expressing a sufficiently large value, such as a 4-byte unsigned integer, and is set to 0 when the system first starts operation.
It shall be cleared and shall return to 0 when it is counted up in the maximum expressible state.

The computer 1 in the second embodiment corresponds to the computer 1 in the first embodiment, and
If the computer 2 or the computer 3 in 1 is made to correspond to the computer 2 in the first embodiment, the operation of the second embodiment is the same as the operation of the first embodiment, and therefore the description thereof is omitted here.

As described above, according to the second embodiment, after updating the record of the data file of the active computer 2, the standby computer 1 is stopped due to a failure or the like until the corresponding record of the standby computer 1 is updated. Even when the standby system computer 1 starts up in the standby system again, the corresponding records in the data files of the computers 1 and 2 of both systems are recognized as inconsistent and are copied from the active system computer 2 to the new standby system computer 1. In addition, since the management unit is on the disk, the contents of the management unit are not erased even if the computers 1 and 2 of both systems are stopped, and the name of the operating system computer at the time of updating is stored. Even after updating with a different value when computer 2 is the active system and computer 1 is the standby system, and then this time is updated with another value when computer 1 is the active system and computer 2 is the standby system, it was run again with the active system + standby system At this time, the records in the data files of the computers 1 and 2 of both systems are recognized as inconsistent and are copied from the active computer 1 to the new standby computer 2. As a result, no matter how the operating mode of the computers 1 and 2 of both systems transits, the data file record in which the contents of the computers 1 and 2 of both systems do not match when the standby system is started up can be correctly recognized, Since only the active computer 1 is copied to the new standby computer 2, the standby startup time is shortened.
Also, when the active-standby system is operating in multiple systems, it is not necessary to synchronize with the data update in the standby computer in the data change process of the active computer, so performance during multiple system operation is not impaired. .

In the first and second embodiments described above, the data file equivalence is explained. However, since the data file naturally contains data, the data file equivalence referred to in the present invention includes the concept of data equivalence. I'm out. Therefore, even in the case of data equivalence, it can be realized by performing the same processing as that described in the first and second embodiments, and in this case, the same effect as that of the first and second embodiments can be obtained.

In the first and second embodiments, the operating computer name is stored in the management unit as the operating computer identification information at the time of updating, but the present invention is not limited to this, and the operating computer number at updating or the operating computer at updating is stored. May store a code indicating
Also in this case, the same effects as those of the first and second embodiments can be obtained.

In the first and second embodiments, the update count counter and the operating computer name at the time of update are stored in the management unit in the record unit of the data file, but these may be stored in the data file unit. Also in this case, the same effects as those of the first and second embodiments can be obtained.

[0033]

As described above, according to the first aspect of the invention, in the operating computer and the standby computer of the dual computer system, after updating the data file, the computers of both systems asynchronously update in record units. Counts up the number of times counter and stores the operating system computer identification information at the time of update, and stores only the records where the standby startup system computer and the operating system computer have different update number counters or the operating system computer identification information at the time of update Since the process of copying from the standby computer to the new standby computer is performed, even if the standby computer stops due to a failure etc. after updating the record of the data file of the operating computer until the corresponding record of the standby computer is updated, When starting up to the standby system, the records in the data files of the computers of both systems are recognized as inconsistent, and the operating system computer sets the new standby system. It is copied to San machines. In addition, since the management unit is on the disk, the contents of the management unit are not erased even if the computers of both systems are stopped, and the operating system computer identification information at the time of updating is stored. After updating when the other computer is the active system and the other computer is the standby system, even if the standby computer is updated to the active system and the active computer is updated to the standby system, the operating system will be updated again. + When operating in the standby system, the records in the data files of the computers in both systems are recognized as inconsistent and are copied from the active computer to the new standby computer. As a result, no matter how the operating mode of the computers on both systems transitions, the data file records with inconsistent contents of the computers on both systems can be correctly recognized when the standby system is started up, and only the inconsistent records can be read from the operating computer. Since it is copied to the new standby system computer, the startup time of the standby system computer is shortened, and when the active system-standby system is operating in a dual system, the standby system computer can change the data of the active system computer. Since it is not necessary to synchronize with the data update of, the effect of not impairing the performance during dual system operation can be obtained.

Further, according to the second invention, in the active computer and the standby computer of the multiple computer system, the first computer
Since it is configured to apply the method equivalent to the invention of the above, even in a multiplexed computer system, after updating the record of the data file of the active computer, the standby computer fails before updating the corresponding record of the standby computer. Even if it is stopped due to such reasons, the corresponding records in the data files of the computers of both systems are recognized as inconsistent at the time of restarting to the standby system, and are copied from the active computer to the new standby computer. Also,
After updating the record of the data file when one computer is the active system and the other computer is the standby system, this time it was updated with another value when the standby computer was the active system and when the active computer was the standby system. Also, when the operating system and the standby system are operated again, the records in the data files of the computers of both systems are recognized as inconsistent and are copied from the active computer to the new standby computer. As a result, no matter how the operating modes of all the system computers are changed, when the standby computer is started up, the data file record in which the contents of the active computer and the new standby computer do not match can be correctly recognized, and the standby startup is started. It is possible to obtain an effect that the time is shortened and the performance during the multi-system operation is not impaired.

[Brief description of drawings]

FIG. 1 is a hardware configuration diagram showing a dual system computer system adopting a data file equivalent processing method according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a data file equivalent processing method according to the first embodiment.

FIG. 3 is a structural diagram of a management unit of the computer 1 according to the first embodiment.

FIG. 4 is a structural diagram of a management unit of the computer 2 in the first embodiment.

FIG. 5 is a flow chart showing a procedure of data change processing of the operating system computer in the first embodiment.

FIG. 6 is a flowchart showing a procedure of data change processing of the standby computer in the first embodiment.

FIG. 7 is a flowchart showing a procedure of a data copy process of the operating system computer in the first embodiment.

FIG. 8 is a flowchart showing a procedure of data copy processing of the new standby computer in the first embodiment.

FIG. 9 is a diagram showing changes in the contents of records in the management unit and data file of each computer in the first embodiment.

FIG. 10 is a diagram showing changes in the contents of records of a management section and a data file of each computer according to the first embodiment.

FIG. 11 is a diagram showing changes in the contents of the records of the management unit and the data file of each computer in the first embodiment.

FIG. 12 is a diagram showing changes in the contents of records of a management section and a data file of each computer in the first embodiment.

FIG. 13 is a diagram showing changes in the contents of the records of the management unit and data file of each computer in the first embodiment.

FIG. 14 is a diagram showing changes in the contents of records of a management section and a data file of each computer in the first embodiment.

FIG. 15 is a diagram showing changes in the contents of records in the management unit and data file of each computer according to the first embodiment.

FIG. 16 is a diagram showing changes in the contents of the records of the management unit and data file of each computer in the first embodiment.

FIG. 17 is a diagram showing changes in the contents of records of a management section and a data file of each computer in the first embodiment.

FIG. 18 is a diagram showing changes in the contents of the records of the management unit and data file of each computer in the first embodiment.

FIG. 19 is a diagram showing changes in the contents of the records of the management unit and data file of each computer in the first embodiment.

FIG. 20 is a diagram showing changes in the contents of records of a management section and a data file of each computer in the first embodiment.

FIG. 21 is a hardware configuration diagram showing a multiplexed computer system adopting the data file equivalent processing method according to the second embodiment of the present invention.

FIG. 22 is a block diagram showing a data file equivalent processing method according to the second embodiment.

FIG. 23 is a structural diagram of a management unit of the computer 1 according to the second embodiment.

FIG. 24 is a structural diagram of a management unit of the computer 2 according to the second embodiment.

FIG. 25 is a structural diagram of a management unit of the computer 3 according to the second embodiment.

FIG. 26 is a hardware configuration diagram showing a two-series computer system adopting a conventional data file equivalent processing method.

FIG. 27 is a block diagram showing a conventional data file equivalent processing method.

FIG. 28 is a flow chart showing the procedure of data change processing of the operating system computer in the conventional data file equivalence processing method.

FIG. 29 is a flowchart showing the procedure of a data change process of the standby computer in the conventional data file equivalence processing method.

FIG. 30 is a flow chart showing the procedure of data copy processing of the operating system computer in the conventional data file equivalence processing method.

FIG. 31 is a flowchart showing a procedure of data copy processing of a new standby system computer in the conventional data file equivalence processing method.

[Explanation of symbols]

1,2,3 Computer, 5 transmission lines, 10, 20, 30
CPU, 11, 21, 31 Main memory, 12, 22, 3
2 disks, 13, 23, 33 transmission I / F, 12
0, 220, 320 management unit, 121, 122, 22
1, 222, 321, 322 data files, 121
1,1221,211,2212,3211,322
1 record, 12110, 12210, 22110,
22210, 32110, 32210 update counter, 12111, 12211, 21211, 2221
1, 32111, 32211, update-time operating system computer name (update-time operating system computer identification information).

Claims (2)

[Claims] 1. Each time two computers connected via a network are used as an active system and the other computer as a standby system for dual operation, each time a data file of an active computer is updated,
In a dual computer system in which update data is equivalent to a data file of a standby computer, each computer is provided with an update counter that counts the number of updates and active computer identification information that identifies the active computer at the time of update. When the equivalence target data file is updated in the operating system computer or standby system computer, the management unit that stores the record unit or the data file unit of each equivalence target data file Update the update counter and the operating system computer identification information at the time of update, and when the computer stopped due to a failure or maintenance starts, and this computer transitions to the standby system, the update counter of both systems or the update Only records or data files with different active computer identification information from the active computer Data file equivalent processing method characterized by copying the standby computer. 2. One of n (n ≧ 3) computers connected via a network is used as an active system, and the other m
In a multi-system computer system in which a large number of units (n-1 ≧ m ≧ 1) are operated as a standby system and each time the data file of the operating computer is updated, the updated data is equivalent to the data file of the standby computer The management unit that stores the update count counter that counts the number of times and the active computer identification information that indicates the identification of the active computer at the time of update in a record unit or data file unit of each equivalence target data file of each computer If the equivalent computer data file is updated in the operating system computer or the standby system computer, the update count counter and the operating system computer identification information at the time of updating the record or the data file are updated, and a failure or maintenance is performed. When the computer that was stopped due to, etc. starts up and this computer transitions to the standby system, the update count of both systems is updated. Data file equivalent processing method characterized in that data or active system computer identification information at the time of update is copied only from the active system computer different records or data files to a new standby computer.