JP2002049502A - Update system for multiprocessor systems - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce resource-relative restrictions and to securely move data by temporarily interrupting service when software of a multiprocessor system is updated, as to a update system for the software of the multiprocessor system and an update system for data. SOLUTION: While the processor of 1+1 redundant constitution of in- operation and standby systems are made to operate independently, a software update part is driven for the processor of the standby state to set new-version software. While the service of the processor of the in-operation system is carried on, a processing event generated by the in-operation system is sent even to the processor of the standby state through an inter-processor communication part. Both the processors perform the same processes to reflect the control state data of the in-operation system even on the standby system gradually. When the variable control state data are all reflected on the standby system a specific time later, the processor of the standby system is switched to in- operation state.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a software update method and a data update method in a multiple processor system.

2. Description of the Related Art In a control system, such as an electronic exchange system, in which downtime is required to be as short as possible, a processor is duplicated, one of which provides a service as an active (act) system, and the other as a standby (standby) system. When the data that is prepared and written to the memory changes from moment to moment, the data is supplied to both the active system and the standby system so that they are synchronously reflected, so that the processors in both systems always hold the same data. This allows switching from the active system to the standby system and from the standby system to the active system at an arbitrary time.

When the software of each processor that performs software control of such a duplicated control system is updated to a new version, the software can be freely updated due to restrictions on resources such as memory size and an increase in processing amount. There is a need for improvements in that it cannot be performed and that it is not possible to cope with a problem with the updated software.

[0004]

2. Description of the Related Art It is required to minimize downtime of an electronic exchange system or the like (mission critical).
In a control system, as a method of improving the reliability of the system, there is a method of duplicating a processing device (processor) for controlling software and using a 1 + 1 redundant exchange used as an operation side (operating processor) / standby side. Yes, the conventional technique will be described.

FIG. 7 is an explanatory diagram of a conventional example, showing an example of a control system of an electronic exchange. In the figure, 80 is a 0-system processor, 81 is a 1-system ($ 1) processor, 82 is a 0-system main memory included in a 0-system ($ 0) processor,
83 is a 1-system main memory included in a 1-system processor, 84 is a switching unit, 85 is a subscriber line control unit for controlling a subscriber's speech path switch (not shown), and 86 is accommodated in the speech path switch. A subscriber terminal 87 such as a telephone is a communication unit between processors.

In the configuration shown in FIG. 7, when online updating (updating such as version upgrade) of system software is performed, one of the processors (including the main memory) 80 is operated by the operating system (including the actual subscriber during operation). While driving with the existing version of software, the new version (upgraded) software is loaded into the main memory 83 of the other standby processor 81 as the control unit driving side), and the operating system is changed to the new version software online. In general, switching to a new version of software is realized by switching to the processor 81 on the side where the information is entered. Thereafter, the processor 80 operating with the existing version software is referred to as an old system, and the processor 81 loaded with the new version software is referred to as a new system.

In this case, in the example of the electronic exchange system,
It is necessary to take over the processing operation to the new system while maintaining the continuity of the call processing service (without disconnecting the call in progress). For this reason, the control data unique to each system existing in the old system (subscriber Data, call state data, etc.) must be taken over to the new system before performing the switching operation. For this reason, the control data in the old main memory 82 is transferred to the new processor 8 by the interprocessor communication unit 87.
Then, the operating system is switched to the new system after the data is reflected on the main memory 83 of the new system, and is converted to a format corresponding to the new version of software on the new system. In this way, online update without call disconnection is generally realized.

[0008]

As described above, after the new system loaded with the new version of software is switched from the standby system to the active system, in order to prevent call disconnection,
Before switching, it is necessary to transfer data that changes every time according to the call status, such as connection information data, to the new system and reflect it synchronously. If the state seems to change, there is a problem that data synchronization between the new and old processors cannot be achieved, and it is necessary to suspend the switching service (call processing stop) in order to prevent a change in the state of the call until the switching is completed. there were.

At this time, in the prior art, since the data in the old main memory itself is transmitted and converted to the new system and converted and taken over to the new system, resources in the new system such as memory size are reduced. However, there is a problem that the degree of freedom of software update is limited due to restrictions such as restrictions on memory address allocation and the like, and an increase in processing on the new system side.

Furthermore, in the prior art, the call processing operation by the new version software is confirmed only at the time of startup on the new system after the switching of the active system, but the operation on the new system is unsatisfactory (NG). In such a case, the old system is switched back, and the call state is reset at that time, so that there is a problem that a call during a call is forcibly disconnected or an overlapping service is interrupted.

In addition, the above problem arises in that data is transferred not only between processors having a 1 + 1 redundancy configuration but also between a plurality of nodes including a plurality of processors (moving data such as terminals handled by each node to a different node). In this case, the data transfer has a problem in the processing, and the same problem occurs.

SUMMARY OF THE INVENTION The present invention solves the above-mentioned problems, and prevents a constraint on resources such as a memory size and an increase in processing without temporarily interrupting a service in a software update in a 1 + 1 redundant configuration processor system. In addition, the present invention provides a software update method in a multi-processor system that can easily switch back to an old system, and a data update method that can realize data transfer in a multi-processor system without interrupting service. The purpose is to:

[0013]

FIG. 1 is a diagram showing a first principle configuration of the present invention. In the figure, 1 is the 0 system (represented by $ 0)
And two systems of 1 system (represented by $ 1), 10 in each processor 1 is a processing device, 11 is a memory storing state data, fixed data, programs, and the like, 12 is control for the control unit 3, An interface control unit (displayed by an IF control unit) for interfacing information such as a state from the control unit 3 between the processing device 10 and the control unit 3, and 13 is a software update unit. Reference numeral 2 denotes an inter-processor communication unit for transferring events and data between processors, and reference numeral 3 denotes control for performing various controls by the processing device 10 based on a program stored in the memory 11 corresponding to the state of many terminals 4. The units 4 are controlled by the control unit 3, and a number of terminals connected by each subscriber line, and 5 is one processor 1 (# 0 or # 0) which is an operating system for controlling the control unit 3 to which each terminal 4 is connected. # 1) represents a switching unit that performs switching connection.

As in the prior art, the software of the standby system is updated to transfer data (for example, call processing data of an exchange) that changes every time in the active system to the standby system (called a new system). In the method that takes over to the new system by format conversion on the new system (new system with updated software),
In view of the fact that it is difficult to avoid interruption of the processing service during the transfer processing, in the present invention, the data in the main memory is not copied / converted and transmitted to the new system but transmitted to the active system (older). By running the same process in both the standby system (new system) and the standby system (new system), the control status of the old system (active system) is gradually reflected in the new system (standby system), and for a certain period of time. After the lapse of time, the data is matched between the old and new processors, and the standby system (new system) is switched as the active system.

In FIG. 1, a control unit 3 normally connected to a large number of terminals 4 is connected to one processor 1, for example, a processor 1 of system 0 ($ 0) as an operating system. The control unit 3 is controlled by the program processing described above, and the processor 1 of the first system (# 1) is separated from the control unit 3 as a standby system and stands by. Update processor software (version upgrade)
In this case, the update state is set by an instruction from an external device (such as a console), the processors 1 of both systems are operated in the independent mode, and the processor 1 of the system 0 is set as the active system by the switching unit 5 by the control unit 3. By driving the software update unit 13 of the processor 1 of the standby system 1 while operating while being connected to the system 1, the software of the processor 1 of the primary system 1 is upgraded, and there is no processing data in the processor 1 of the primary system 1. State. Thereafter, the fixed data (for example, attribute data of the terminal 4) in the memory 11 of the processor 0 of the active system is transferred to the processor 1 of the standby system 1 via the interprocessor communication unit 2. The data is expanded in the memory 11 in the processor 1 of the first system, and the first system converts the data into a format corresponding to the new software and stores it.

After the fixed data of system 0 is expanded to system 1, event data (variable data) generated by control operation of terminal 4 of control unit 3 in processor 1 of system 0 of operation system
Is also transferred to the first system, the processor 1 of the first system performs processing with new software, generates corresponding state data, and updates it by storing it in the memory 11. No control processing is performed because there is no data. If an error occurs due to processing by the upgraded software on the first system, the update is stopped and the original software is restored. When a certain period of time elapses after the processing of the first system processor is normally executed, the processing data (variable data) becomes only the data relating to the control executed after the start of the update, and the contents of the memory are changed between the active system and the standby system. When they completely match, the switching unit 5 is driven to switch the first processor 1 to the active processor. At this time,
Since the status data of the control processing matches the active system, the processing by the software updated by switching requires only a short interruption time, and if no abnormality occurs due to the updated software, the software of the 0-system processor that became the standby system Is also updated to match the software on both processors.

FIG. 2 shows a second principle configuration of the present invention. The configuration shown in FIG. 2 applies the principle of the invention shown in FIG. 1 to data transfer in a system composed of a plurality of processors. In FIG. 2, reference numerals 1a and 1b denote processors that perform processing for different control units 3a and 3b, and 10, 11, and 12 in each processor 1a and 1b are the same as those of the same reference numerals in FIG. The transfer unit, 2 is an inter-processor communication unit, and 3a, 3b are terminals 4a,
The control units 4a and 4b for controlling the 4b are connected to the control units 3a and 3b, respectively, and the processors 1a and 1b
Are controlled by the processor 1a at first, and the data of the terminal is transferred to the control processor 1b and the control unit 3b controls the control unit 3b.
The terminal whose connection is changed to b, and 5 is a switching unit.

In FIG. 2, usually, a plurality of processors 1
a and 1b are control units 3 each of which is an individual control object.
a and 3b connected to terminals 4a and 4b,
When the terminal 4c is controlled by the processor 1a via the control unit 3a, fixed data and status data relating to the terminal 4c are stored in the memory 11 of the processor 1a. At this time, if the terminal 4c needs to be relocated from the subordinate of the processor 1a to the subordinate of the processor 1b due to movement or the like, the data relocation of the relocation destination processor 1b is performed according to a data relocation instruction from the outside (maintenance console or the like). The unit 14 is driven to generate a data transfer notice, which is notified to the interface control unit 12 of the transfer destination processor 1b and the transfer source processor 1a, and the fixed (operating) data of each data transfer target terminal 4c. Set the status indicating that the relocation is in progress.

Thus, the fixed data relating to the terminal 4c to be relocated is transferred from the memory 11 of the relocation source to the memory 11 of the processor 1b of the relocation destination via the inter-processor communication section 2, and the transfer of the fixed data to the relocation destination is completed. Then, the processor 1a of the relocation originates the subscriber for the data relocation.
The processing event occurring thereafter is sent to the own processor 1a.
At the same time, the data is sent to the transfer destination processor 1b by the interface control unit 12 through the inter-processor communication unit 2. The transfer destination processor 1b performs the same processing as that of the transfer source processor 1a on this processing event.
However, the control data such as a response message generated in the processor 1a at the relocation destination is discarded, and the control of the subscriber or the like is not performed. After this state continues, the transfer destination processor 1
When the operation stability at step b is confirmed, the state of the transfer source and the transfer destination processor regarding the transfer subscriber during the transfer is released, and the subscriber is connected to the transfer destination processor 1b side,
Thereafter, the data of this subscriber remaining in the memory 11 of the transfer source is deleted.

[0020]

FIG. 3 shows the structure of the first embodiment. The first embodiment corresponds to the principle configuration of FIG. 1 and is an example applied to an exchange for performing call processing and connection for a large number of subscriber terminals. In FIG. 3, reference numerals 1, 10 to 13, and 2 to 4 correspond to the respective units or units having the same reference numerals in FIG. 1, and 1-0 denotes an old processor operating as an active system.
1-1 is a new system side processor that operates as a standby system, 10
0-0 and 10-1 manage the subscriber status according to the call processing events received from the subscriber interface control unit and the state of the system, and generate and issue necessary call processing events. 11-1 is a storage device, 11a in each storage device is status data, 11b is a program, and 11c is fixed data. Reference numerals 12-0 and 12-1 denote subscriber interface control units. The subscriber detects a state change such as call origination / end of call, notifies the call processing control unit, and according to the instruction of the call processing control unit. Controls the other device and receives notification of the start and end of the software updater from the software update scenario control unit.
A function to transmit call processing events to the new system side in parallel during software update, and to notify the call processing events received from the old system side to the call processing control unit on the new system side and process response events from the call processing control unit. I do.

Reference numerals 13-0 and 13-1 denote software update scenario control units, which manage and control each functional unit in the system according to the software update scenario according to the progress status of the software update and the system state. It notifies the interface controller of the start / end of the software update.
14-0 and 14-1 are data transfer / conversion units for performing data transfer using the inter-processor communication unit and data conversion on the new system side. 2 is an inter-processor communication unit, 3 is a subscriber line control unit corresponding to the control unit 3 of FIG. 1, and 4 is a terminal 4 of FIG.
Is a subscriber terminal corresponding to.

The processing flow of the software update in the configuration of FIG. 3 will be described with reference to FIGS. 4 and 5 show the timing relations (No. 1) and (No. 2) of the software update processing, and (1), (2),.
The processing corresponding to each number will be described below.

(1) The processor 1-0 is active (old),
When the processor 1-1 is used as a standby system (new system) and actual control of the subscriber terminal 4 and the trunk (not shown) is performed in the old system, software is executed in the target system by a maintenance person such as an operator console. Declare transition to update state. In response to this, the software update scenario control unit 13-0 shifts the processor from the duplex mode to a mode in which processing is performed independently of each other, and operates the standby processor in parallel with the operating processor running. The new version of the system software is loaded into the storage device 11-0, and the standby system side program is initialized.

(2) Thereafter, the software update scenario control unit 13 on the standby system (new system) side using the inter-processor communication unit 2 from the active system (old system) processor 1-0.
-1 is notified of the start of update. On the standby system (new system) side, an update start notification is notified from the software update scenario control unit 13-1 to the subscriber interface control unit 12-1.

(3) Subscriber interface control unit 12-1
Then, a flag indicating "updating" is set.

(4) The software update scenario control unit 13-0 uses the inter-processor communication unit 2 and the data transfer / conversion unit 14-0 to use the operation data (subscriber attribute) registered in the system at that time. Data such as data that does not depend on the call state is transmitted from the active system to the standby system.

(5) The standby system (new system) converts the transmitted data into the self-storage device 11-1 while converting the data into a format corresponding to the new software as necessary.
As a result, the operation data is synchronized between the active system and the standby system.

(6) Upon completion of the deployment of the operation data in (5), the software update scenario control unit 13-0 on the active (old) side starts updating to the subscriber interface control unit 12-0 on its own. Notify.

(7) Upon receiving the update start notification, the subscriber interface control unit 12-0 “updates”
Is set.

(8) On the active system (old system) side, a call processing event relating to a call newly generated thereafter is passed to the call processing control unit 10-0 of the own system to perform call processing.
It flows to the subscriber interface control unit 12-1 on the standby side using the inter-processor communication unit 2.

(9) When the flag indicating “updating” is set, the subscriber interface control unit 12-1 on the standby side transmits the call processing event from the active side to the normal local side. The flow is sent to the call processing control unit 10-1 to perform the same call processing as that of the active system, and call state data equivalent to that of the active system is generated on the main memory of the standby system. At this time, the call processing control unit 10 on the standby system side only performs the call processing only by the memory operation and constructs the data equivalent to the current call state data on its own storage device 11-0. -0, a call processing event such as a response message is generated by the subscriber interface controller 1 on the standby side.
Control is performed using the above-mentioned "updating" flag so that discarding is not performed in 2-1 and hardware such as a subscriber line or traffic is not controlled.

(10) If it is necessary to cancel the update work and return to the old version of the software because of a processing error in the software of the standby processor 1-1, a notification of an abnormality from the standby system to the active system The active processor 1-0 detects this by monitoring the standby system status, and the active processor 1-0 operates as it is, disconnects the standby processor 1-1 and cancels the update operation, thereby restoring the update. It is possible and does not affect the service at all.

(11) In this state, when a sufficient time has elapsed, all the calls made before the start of the software update are terminated, and the call control data relating to only the call made after the start of the software update for both the new and old systems is lost. When the state is considered to exist, the processor 1-1 in the standby system (new system) is switched to the active system. The service is temporarily interrupted only at this time because it is necessary to run the initial settings here. However, the new system has already established the call state data equivalent to the active system. By doing so, initial setting without call disconnection becomes possible. In addition, since the service is not temporarily interrupted even during the data conversion, the service interruption time in the entire work can be relatively short.

Thereafter, subscriber control is performed only on the new system side, and
If it is confirmed that there is no abnormal operation during these operation procedures, the old software remaining in the old operating system is erased, and both processors are unified to the new software, the updating state is released, and the process ends.

FIG. 6 shows the configuration of the second embodiment. The second embodiment corresponds to the second principle configuration of the present invention shown in FIG. 2 and is an example applied to data transfer in a plurality of node systems.

In FIG. 6, reference numerals 1a and 1b denote a transfer source processor (or node) and a transfer destination processor (or node), respectively. In this embodiment, two processors (nodes) are provided. The same applies even if more than one processor (node) is provided. 10a and 10b in each processor are the same call processing control units as 10 in FIG. 3, and 11a and 11b are 11 in FIG.
-0 and 11-1 are storage devices similar to the above, where 110 is status data, 111 is a program, 112 is fixed data, and 12a and 12b are 12- in FIG.
0, 12-1 is the same subscriber interface control unit. In the second embodiment, the data transfer scenario control unit (15a, 15b, which will be described later) receives a notification of the start and end of the data transfer. A function for transmitting a call processing event in parallel to the transfer destination processor side, and a call processing event received from the transfer source processor side on the transfer destination processor side is notified to the call processing control sections 10a and 10b, and a response event from the call control section is notified. To process. 14a and 14b correspond to FIG.
And a data transfer / conversion unit similar to 14-0 and 14-1 in, and particularly performs data conversion in the relocation destination processor. Reference numerals 15a and 15b denote data transfer scenario controllers for managing and controlling the respective functional units in the system according to the data transfer scenario according to the progress of data transfer and the system status, and in particular, the subscriber interface controllers 12a and 12b.
Notification of the start / end of the data transfer to b. Reference numeral 2 denotes an inter-processor communication unit, 3a and 3b denote subscriber line control units managed and controlled by the processors 1a and 1b, respectively, and 4 denotes a subscriber terminal.

The operation of the second embodiment shown in FIG. 6 will be described below in the order of operation. In this example, the subscriber terminal 4 accommodated in the subscriber control unit 3a under the control of the processor 1a.
Is transferred to the subscriber control unit 3b under the control of the processor 1b.

(1) The transition to the data transfer state is declared in the transfer destination system (processor 1b) by a maintenance person intervention means (not shown) such as a console. In the transfer destination processor, a data transfer start notification is sent from the data transfer scenario control unit 15b to the subscriber interface control unit 12b, and the subscriber interface control unit 12b sets a flag indicating "migrating" for the data transfer target subscriber. .

(2) Transfer operation data (fixed data that does not depend on the call state, such as subscriber attribute data) relating to the data transfer target subscriber. As a general means, it is possible to re-register with the relocation destination system using an offline command or the like. Alternatively, the data may be automatically transferred to the destination using the inter-processor communication unit 2 or the data transfer / conversion unit 14a.

(3) When the transfer of the operation data according to the above (2) is completed, the transfer source system declares the shift to the data transfer state. This declaration can be made from the console by the maintenance person, but can also be notified from the relocation destination using the inter-processor communication unit 2. Relocation source processor 1
In a, the data transfer scenario control unit 15a notifies the start of the update to the subscriber interface control unit 12a of the own system, and the subscriber interface control unit 12a sets a flag indicating "migrating" for the data transfer target subscriber. On the source processor side, the call processing events for calls that have occurred since then for the data transfer target subscriber are:
At the same time, the call is sent to the call processing control unit 10a of the own processor to perform the call processing, and at the same time, the call is sent to the subscriber interface control unit 12a on the transfer destination processor side using the inter-processor communication unit 2.

(4) In this state, if operation stability is confirmed in the relocation destination processor after operation for a while, the "transferring" flag is turned off together with the relocation source processor and the relocation destination processor, and the subscriber connection is switched to the relocation destination processor. Switch to the side.

(5) When it is confirmed that there is no abnormal operation in the above procedure, the data to be relocated remaining in the old operating system is deleted, the "relocating" state is released, and the process ends.

According to the second embodiment, when the subscriber accommodation is changed between processors (nodes), the operation in the new accommodation node can be verified to some extent in advance while the call processing (service) of the old accommodation node is continued. Thus, when the processing on the new accommodation node side fails, the possibility of service interruption for the relocation target subscriber can be reduced.

(Supplementary Note 1) In a multiple processor system, the software update unit is driven for the standby processor to set new version software in a state where the active and standby 1 + 1 redundant processors are operated independently. ,
Processing events that occur in the active system are also sent to the standby processor via the inter-processor communication unit while the service of the active processor is continued, and the processing status of the active system is reduced by performing the same processing in both processors. A software update method in which a standby processor is switched to an active system when the variable control state data is reflected in the standby system after a lapse of a predetermined time and is gradually reflected in the standby system.

(Supplementary note 2) In Supplementary note 1, an event to be processed in the active system is flowed to the standby processor while the service of the active processor is continued, and the control state data is gradually reflected in the standby system. A software update method in which the operation of the standby processor is verified, and if the processing of the standby processor becomes abnormal, the switching of the standby processor to the active system is prohibited.

(Supplementary note 3) In Supplementary note 1, a flag is provided in the active processor to indicate that data to the standby system is being taken over online, and based on the flag, it is necessary to transmit a processing event to the standby system. A software update method characterized by judging and controlling the software.

(Supplementary note 4) In Supplementary note 1, a flag is provided to the standby processor to indicate that data is being taken over online, and it is determined based on the flag whether the output of the event processing result of the own processor is necessary. Software update method characterized by controlling

(Supplementary Note 5) In Supplementary Note 1, the active processor transfers fixed data such as the attribute of the terminal under management to the standby processor, and then processes a processing event generated in the active system by the standby system. Software update method characterized by streaming to the processor.

(Supplementary Note 6) In a system with a plurality of processors, each processor holds data of terminals and the like housed and managed by each processor, and when a terminal or the like housed in one processor is moved to another processor, The data to be relocated is transferred from the relocation source processor to the relocation destination processor, and the processing event generated by continuing the service of the relocation source processor is also sent to the relocation destination processor via the inter-processor communication unit. By performing the same processing, the control state data of the transfer source is gradually reflected in the transfer destination. When a predetermined time has elapsed and the variable control state data has been completely reflected in the transfer destination, the processor of the transfer destination executes the processing. A data update method for switching management of the terminal to be relocated.

[0050]

According to the present invention, the following effects can be obtained.

When updating software in a multiprocessor system having a 1 + 1 redundancy configuration,
Since there is no need to send and convert call control data between the active (old) and standby (new) storage devices, there is no service interruption during that time, and the interruption time is minimized only during system switching. Can be.

In the software update of the 1 + 1 redundant configuration, when the operation with the new file is NG, the software update is disabled and the operation with the old system file is continued, but before the switching (while the old system is operating as the active system). The operation in the new system can be confirmed to some extent during the independent operation of the two systems), so that the probability of the new system operation being NG after switching can be reduced, and the probability of service interruption due to this can be reduced.

In the software update of the 1 + 1 redundant configuration, takeover data is generated by processing on the new system side (processing of new version software). Therefore, restrictions on resources such as memory size on the new system side and allocation of memory addresses are performed. Therefore, the degree of freedom of the software update operation can be improved without being restricted by the restrictions such as the above and the processing increase on the new system side.

In a multi-node configuration, even when data is transferred between processors (nodes), the operation at the transfer destination is verified in advance while the process (service) of the transfer source is continued, so that the data at the transfer destination can be verified. The processing can be ensured, and the possibility of service interruption for the relocation target subscriber can be reduced even if the processing fails.

[Brief description of the drawings]

FIG. 1 is a diagram showing a first principle configuration of the present invention.

FIG. 2 is a diagram showing a second principle configuration of the present invention.

FIG. 3 shows a configuration of the first embodiment.

FIG. 4 is a diagram showing a timing relationship (part 1) of a software update process.

FIG. 5 is a diagram showing a timing relationship (part 2) of a software update process.

FIG. 6 shows a configuration of a second embodiment.

FIG. 7 is an explanatory diagram of a conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 0-system ($ 0) and 1-system ($ 1) processors 10 processing unit 11 memory 12 interface (IF) control unit 13 software update unit 2 interprocessor communication unit 3 control unit 4 terminal 5 switching unit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04M 3/22 H04Q 3/545 5K026 H04Q 3/545 G06F 9/06 630D 5K051 F term (Reference) 5B018 GA04 HA05 MA12 PA03 QA01 5B034 BB02 BB17 CC01 5B076 CA06 EA11 EA17 5B082 AA02 DE07 GA14 GB06 HA03 5K019 AA08 BA24 CA05 CC05 EA28 EA29 EA32.

Claims (3)

[Claims] In a multiple processor system, a software update unit is driven for a standby processor to set a new version of software in a state in which each processor of a 1 + 1 redundant configuration of an active system and a standby system is operated independently. Process events that occur in the active system are also sent to the standby processor via the inter-processor communication unit while the service of the active processor is continued, and the same processing is performed by both processors to wait for the control status data of the active system. A software update method in which the standby processor is switched to the active system when the variable control state data is gradually reflected in the standby system after a predetermined period of time and all the variable control state data is reflected in the standby system. 2. The method according to claim 1, wherein an event to be processed in the active system is sent to the standby processor while the service of the active processor is continued, and the control state data is gradually reflected in the standby system. A software update method for verifying the operation of a standby processor and prohibiting switching of the standby processor to an active system when processing of the standby processor becomes abnormal. 3. In a multiprocessor system, when each processor holds data of a terminal or the like housed and managed by each processor, and when a terminal or the like housed in one processor is relocated to another processor, the source of the relocation is The target data is transferred from the processor to the destination processor, the service of the source processor is continued, and the processing events that occur are sent to the destination processor via the inter-processor communication unit, and the two processors perform the same processing. And the control state data of the transfer source is gradually reflected in the transfer destination. When the variable control state data is completely reflected in the transfer destination after a lapse of a predetermined time, the processor of the transfer destination performs the above-mentioned process. A method of updating data, characterized by switching the management of terminals.