High-reliability redundant backup calculation method and system based on multi-activity and multi-control

Abstract

The invention relates to the technical field of redundant backup calculation, in particular to a high-reliability redundant backup calculation method and system based on multi-activity and multi-control. According to the scheme, the front plug-in module comprises the two high-performance processors and the first data exchange unit, the first processor and the second processor are backups for each other, and when the first processor has an operation fault, the second processor establishes communication connection with the first device through the first data exchange unit; the communication connection between the first equipment and the first processor is disconnected, so that the second processor is seamlessly connected with the calculation task of the first equipment, and the data calculation interruption duration caused by processor faults is greatly reduced; namely, the first processor and the second processor in the scheme are redundant backups for each other, so that the overall reliability of the server can be improved on the premise of ensuring the computing power.

Description

Technical Field

[0001] The present invention relates to the field of redundant backup computing technology, and in particular to a highly reliable redundant backup computing method and system based on multi-active and multi-control. Background Art

[0002] The ATCA standard (Advanced Telecom Computing Architecture) is a new generation of mainstream industrial computing technology that is widely used in telecommunications, aerospace, industrial control, medical equipment, intelligent transportation, military equipment and other fields. The ATCA standard provides a modular, compatible and scalable hardware architecture for the next generation of converged communications and data network applications.

[0003] At present, the reliability of server chassis using the ATCA standard is achieved only by the stability of a single circuit, without any additional redundant backup design, resulting in the reliability of server operation to be improved. Summary of the invention

[0004] The main purpose of the present invention is to provide a high-reliability redundant backup calculation method and system based on multi-active and multi-control, aiming to solve the problem that the reliability of a server chassis using the ATCA standard is only achieved by relying on the stability of a single circuit, resulting in the reliability of the server operation to be improved.

[0005] The technical solution proposed by the present invention is: A high-reliability redundant backup computing method based on multi-active and multi-control is applied to a high-reliability redundant backup computing system based on multi-active and multi-control; the system includes a front-end module; the front-end module includes a first processor, a second processor and a first data exchange unit; the first processor and the second processor are both communicatively connected to the first data exchange unit; the first data exchange unit is communicatively connected to a service device; the method includes: When the front plug-in module is started and meets the first preset condition, the front plug-in module controls the first processor to be in a running state and controls the second processor to be in a standby state, wherein the first preset condition is that both the first processor and the second processor can operate normally; The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, marks the service device connected to the first processor as the first device, and calculates the service data sent by the first device to implement the service function; When the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor; The second processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0006] Preferably, the system further comprises a rear-insertion module and a connection backplane; the front-insertion module further comprises a first storage unit; the first processor and the second processor are both communicatively connected to the first storage unit; the rear-insertion module comprises a third processor, a fourth processor, a second storage unit and a second data exchange unit; the third processor and the fourth processor are both communicatively connected to the second data exchange unit; the third processor and the fourth processor are both communicatively connected to the second storage unit; the second data exchange unit is communicatively connected to a service device; the front-insertion module and the rear-insertion module are both communicatively connected to the connection backplane; the method further comprises: After the rear module is started, the third processor and the fourth processor are controlled to be in a standby state; The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, marks the service device connected to the first processor as a first device, and calculates the service data sent by the first device to implement the service function, and then further includes: The first processor calculates the business data to obtain result data and stores it in the first storage unit; The front-insertion module sends the result data in the first storage unit to the second storage unit for backup storage via the connection backplane; When both the first processor and the second processor have operation failures, the third processor establishes an NTB channel through the second data exchange unit, the connection backplane and the first data exchange unit; The third processor establishes a communication connection with the first device based on the NTB channel, and disconnects the communication connection between the first device and the first processor; The third processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0007] Preferably, the front-end module further includes a first reset circuit unit; the first processor and the second processor are both communicatively connected to the first reset circuit unit; when the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor, and then further includes: The second processor sends a reset command to the first reset circuit unit; The first reset circuit unit restarts the first processor based on the reset command; The front-end module controls the first processor to be in a standby state after restart.

[0008] Preferably, the front-end module controls the first processor to be in a standby state after restart, and then further includes: When the second processor fails to operate, the first processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the second processor; The first processor obtains the service data sent by the first device, and calculates the service data to implement the service function; The first processor sends a reset command to the first reset circuit unit; The first reset circuit unit restarts the second processor based on the reset command; The front-end module controls the second processor to be in a standby state after restart.

[0009] Preferably, it also includes: When the front plug-in module is started and the second preset condition is met, the front plug-in module controls the first processor to be in a running state and marks the second processor as a faulty processor, wherein the second preset condition is: the first processor can operate normally, and the second processor cannot operate normally; The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, and calculates the service data to implement the service function; When the front plug-in module is started and the third preset condition is met, the front plug-in module controls the second processor to be in a running state and marks the first processor as a faulty processor, wherein the third preset condition is: the second processor can operate normally, and the first processor cannot operate normally; The second processor establishes a communication connection with the service device through the first data exchange unit to obtain the service data sent by the service device, and calculates the service data to realize the service function.

[0010] Preferably, the rear plug-in module further includes a second reset circuit unit; the third processor and the fourth processor are both communicatively connected to the second reset circuit unit; the third processor obtains the service data sent by the first device, and calculates the service data to implement the service function, and then further includes: The third processor sends a reset command to the second reset circuit unit; The second reset circuit unit sends the reset command to the first reset circuit unit via the connection backplane; The first reset circuit unit restarts the first processor and the second processor based on the reset command; The front-insertion module controls the first processor and the second processor to be in a standby state after restart; The first processor re-establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the third processor; The first processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0011] Preferably, the third processor obtains the service data sent by the first device, and calculates the service data to implement the service function, and then further includes: When the third processor fails to operate, the fourth processor establishes a communication connection with the first device based on the NTB channel, and disconnects the communication connection between the first device and the third processor; The fourth processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0012] Preferably, after the rear plug-in module is started, the third processor and the fourth processor are controlled to be in a standby state, and then the method further includes: The third processor establishes a communication connection with the service device through the second data exchange unit to obtain service data sent by the service device, marks the service device connected to the third processor as a second device, and calculates the service data sent by the second device to implement the service function; When the third processor fails to operate, the fourth processor establishes a communication connection with the second device through the second data exchange unit, and disconnects the communication connection between the second device and the third processor; The fourth processor obtains the service data sent by the second device, and calculates the service data to implement the service function; The fourth processor sends a reset command to the second reset circuit unit; The second reset circuit unit restarts the third processor based on the reset command; The rear module controls the third processor to be in a standby state after the restart.

[0013] Preferably, the first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, marks the service device communicatively connected to the first processor as the first device, and calculates the service data sent by the first device to implement the service function, and then further includes: The front-end module obtains operating status parameters of the first processor, wherein the operating status parameters include occupancy rate, operating frequency and operating temperature; The front-end module obtains the average occupancy rate, average operating frequency and average operating temperature of the first processor within a preset time period in the past based on the operating status parameters; When the operating state parameter of the first processor meets the fourth preset condition, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor, wherein the fourth preset condition is: the average occupancy rate within the past preset time period is greater than the first preset occupancy rate, the average operating frequency within the past preset time period is greater than the preset frequency, and the average operating temperature within the past preset time period is greater than the preset temperature; The second processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0014] The present invention also proposes a high-reliability redundant backup computing system based on multi-active and multi-control, which applies a high-reliability redundant backup computing method based on multi-active and multi-control; the system includes a front-end module; the front-end module includes a first processor, a second processor and a first data exchange unit; the first processor and the second processor are both communicatively connected to the first data exchange unit; the first data exchange unit is communicatively connected to a business device.

[0015] Through the above technical solution, the following beneficial effects can be achieved: The high-reliability redundant backup computing method based on multi-active and multi-control proposed in the present invention can improve the reliability of server operation; this solution is based on the ATCA-E architecture. On the basis of the ATCA-E architecture, the front plug-in module (i.e., the front plug-in board) includes two high-performance processors (a first processor and a second processor) and a first data exchange unit. The first processor and the second processor back up each other. When the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit (i.e., the second processor immediately takes over the first device), and disconnects the communication connection between the first device and the first processor, so that the second processor seamlessly connects the computing tasks for the first device, greatly reducing the data computing interruption time caused by the processor failure; that is, the first processor and the second processor in this solution are redundant backups for each other, which can improve the overall reliability of the server on the premise of ensuring computing power. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0017] Figure 1 A flowchart of the steps of a first embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-control proposed by the present invention; Figure 2A schematic diagram of a module structure diagram of an embodiment of a high-reliability redundant backup computing system based on multi-active and multi-controllers proposed by the present invention; Figure 3 This is a schematic diagram of the module structure of another embodiment of a high-reliability redundant backup computing system based on multi-active and multi-controllers proposed by the present invention. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0019] The present invention proposes a high-reliability redundant backup calculation method and system based on multi-active and multi-control.

[0020] As attached Figure 1 -Attached Figure 3 As shown, in a first embodiment of a high-reliability redundant backup computing method based on multi-active and multi-control proposed by the present invention, the high-reliability redundant backup computing method based on multi-active and multi-control is applied to a high-reliability redundant backup computing system based on multi-active and multi-control; the system includes a front-insertion module; the front-insertion module includes a first processor, a second processor and a first data exchange unit (PCIe-SWITCH in this embodiment, PCIe-Switch is PCI Express Switch, which is a hardware device for expanding and managing PCIe bus); the first processor and the second processor are both connected to the first data exchange unit through PCIe5.0 bus communication; the first data exchange unit is connected to a business device through PCIe5.0 bus communication; this embodiment includes the following steps: Step S110: When the front-end module is started and satisfies a first preset condition, the front-end module controls the first processor to be in a running state and controls the second processor to be in a standby state, wherein the first preset condition is that both the first processor and the second processor can operate normally.

[0021] Specifically, the front plug-in module in this embodiment is a front plug-in board of the ATCA-E chassis architecture. The ATCA-E chassis architecture expands the size of the rear plug-in board based on the ATCA chassis architecture, so that the size of the rear plug-in board is consistent with the size of the front plug-in board, so that the rear plug-in board has the size conditions for being a computing node. The first processor and the second processor in this embodiment are both Feiteng Tengyun 5000C, specifically 5000C-64 multi-core processors; the business equipment here is the hardware equipment used by different customers in actual applications to achieve different computing needs (such as a host device for image recognition analysis).

[0022] Step S120: The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, marks the service device communicatively connected to the first processor as a first device, and calculates the service data sent by the first device to implement service functions.

[0023] Specifically, when both the first processor and the second processor can operate normally, the first processor is operated preferentially and the second processor is set to standby mode. Thereafter, the first processor establishes a communication connection with the service device and calculates the service data sent to the first device to implement the service function.

[0024] Step S130: When the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor.

[0025] Step S140: The second processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0026] Specifically, when the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit (i.e., the second processor immediately takes over the first device), and disconnects the communication connection between the first device and the first processor, and uses the second processor to seamlessly connect the computing tasks for the first device.

[0027] The high-reliability redundant backup computing method based on multi-active and multi-control proposed in the present invention can improve the reliability of server operation; this solution is based on the ATCA-E architecture. On the basis of the ATCA-E architecture, the front plug-in module (i.e., the front plug-in board) includes two high-performance processors (a first processor and a second processor) and a first data exchange unit. The first processor and the second processor back up each other. When the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit (i.e., the second processor immediately takes over the first device), and disconnects the communication connection between the first device and the first processor, so that the second processor seamlessly connects the computing tasks for the first device, greatly reducing the data computing interruption time caused by the processor failure; that is, the first processor and the second processor in this solution are redundant backups for each other, which can improve the overall reliability of the server on the premise of ensuring computing power.

[0028] In a second embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-control proposed by the present invention, based on the first embodiment, the system also includes a rear plug-in module (i.e., a rear plug-in board) and a connecting backplane; the front plug-in module also includes a first storage unit; the first processor and the second processor are both communicatively connected to the first storage unit; the rear plug-in module includes a third processor, a fourth processor, a second storage unit and a second data exchange unit (also a PCIe-SWITCH); the third processor and the fourth processor are both communicatively connected to the second data exchange unit; the third processor and the fourth processor are both communicatively connected to the second storage unit; the second data exchange unit is communicatively connected to a service device; the front plug-in module and the rear plug-in module are both communicatively connected to the connecting backplane; the models of the third processor and the fourth processor are consistent with the model of the first processor; this embodiment also includes the following steps: Step S210: After the rear module is started, the third processor and the fourth processor are controlled to be in a standby state.

[0029] Specifically, after the rear-insertion module is started, the third processor and the fourth processor are both in a standby state, and the rear-insertion module is used to perform redundant backup for the front-insertion module.

[0030] Step S120 further includes the following steps: Step S220: The first processor calculates the business data to obtain result data and stores it in the first storage unit.

[0031] Step S230: the front-end module sends the result data in the first storage unit to the second storage unit through the connection backplane for backup storage.

[0032] Specifically, during normal operation, the front module serves as an active computing node to meet the functions of the business device, and the rear module serves as a data backup module (sending the result data in the first storage unit to the second storage unit for backup storage by connecting to the backplane).

[0033] Step S240: When both the first processor and the second processor have operation failures, the third processor establishes an NTB channel through the second data exchange unit, the connection backplane and the first data exchange unit.

[0034] Specifically, the NTB channel is a point-to-point connection channel based on PCIe-SWITCH, which is used to connect two systems with independent I / O (input / output) and memory domains; in the NTB channel, each NTB endpoint has one or more windows exposed to the local system, and write operations to these windows are mirrored to the memory of the remote system; because it is a direct connection based on the PCIe link, the data transmission path is short, and the NTB channel can significantly reduce data transmission delays compared to transmission through the CPU or through the network protocol stack.

[0035] Step S250: the third processor establishes a communication connection with the first device based on the NTB channel, and disconnects the communication connection between the first device and the first processor.

[0036] Step S260: The third processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0037] Specifically, when both the first processor and the second processor experience operational failures, the entire front-end module cannot continue to perform computing tasks, so the third processor immediately takes over the first device to minimize adverse effects on the business device.

[0038] At present, in the traditional ATCA chassis and CPCI chassis design, only the front-inserted module can be used as a computing node, and its rear-inserted module can only be used as an interface expansion due to size reasons. Its reliability design basically relies on the stability of a single circuit, and there is no additional redundant backup design. This solution is based on the ATCA-E architecture. On the basis of the ATCA architecture, the ATCA-E architecture expands the size of the rear-inserted module, so that the size of the rear-inserted module is consistent with the size of the front-inserted module, so that the rear-inserted module has the size conditions as a computing node, so that the rear-inserted module can also be used as a computing node to perform computing tasks on business equipment; the present invention uses a plurality of processors (first processor, second processor, third processor and fourth processor) on the ATCA-E architecture as a redundant backup method, thereby improving the overall reliability of the server while ensuring high computing power.

[0039] In a third embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-control proposed by the present invention, based on the first embodiment, the front-end module further includes a first reset circuit unit; the first processor and the second processor are both communicatively connected to the first reset circuit unit; step S130, and then further includes the following steps: Step S310: the second processor sends a reset command to the first reset circuit unit.

[0040] Step S320: the first reset circuit unit restarts the first processor based on the reset command.

[0041] Specifically, the first reset circuit unit generates a reset signal based on the reset command, and restarts the first processor based on the reset signal.

[0042] Step S330: the front-end module controls the first processor to be in a standby state after restart.

[0043] Specifically, when the second processor takes over the first device and implements the business function, it immediately sends a reset command to the first reset circuit unit. The first reset circuit unit restarts the first processor based on the reset command, and the first processor after the restart is in a standby state, that is, the first processor in the standby state can provide redundant backup for the working second processor again.

[0044] In a fourth embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-controller systems proposed by the present invention, based on the third embodiment, step S330 further includes the following steps: Step S410: When the second processor fails to operate, the first processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the second processor.

[0045] Step S420: The first processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0046] Step S430: the first processor sends a reset command to the first reset circuit unit.

[0047] Step S440: the first reset circuit unit restarts the second processor based on the reset command.

[0048] Specifically, the first reset circuit unit generates a reset signal based on the reset command, and restarts the second processor based on the reset signal.

[0049] Step S450: the front-end module controls the second processor to be in a standby state after restart.

[0050] Specifically, the first processor is in a standby state after restart, that is, the first processor in the standby state can provide redundant backup for the working second processor again; when the second processor fails, the first processor takes over the first device again and restarts the first processor, so that the second processor can provide redundant backup for the working first processor again.

[0051] In a fifth embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-control proposed by the present invention, based on the first embodiment, this embodiment further includes the following steps: Step S510: When the front module is started and meets the second preset condition, the front module controls the first processor to be in operation and marks the second processor as a faulty processor, wherein the second preset condition is: the first processor can operate normally, and the second processor cannot operate normally.

[0052] Step S520: The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, and calculates the service data to implement the service function.

[0053] Specifically, when the front-end module is started and only the first processor can operate normally, the first processor is operated to calculate the service data sent by the service device to implement the service function.

[0054] Step S530: When the front module is started and satisfies the third preset condition, the front module controls the second processor to be in operation and marks the first processor as a faulty processor, wherein the third preset condition is that the second processor can operate normally and the first processor cannot operate normally.

[0055] Step S540: The second processor establishes a communication connection with the service device through the first data exchange unit to obtain the service data sent by the service device, and calculates the service data to implement the service function.

[0056] Specifically, when the front-end module is started and only the second processor can operate normally, the second processor is operated to calculate the service data sent by the service device to implement the service function.

[0057] In a sixth embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-control proposed by the present invention, based on the second embodiment, the rear plug-in module further includes a second reset circuit unit; the third processor and the fourth processor are both communicatively connected to the second reset circuit unit; step S260, and then further includes the following steps: Step S610: the third processor sends a reset command to the second reset circuit unit.

[0058] Step S630: the second reset circuit unit sends the reset command to the first reset circuit unit through the connection backplane.

[0059] Step S640: The first reset circuit unit restarts the first processor and the second processor based on the reset command.

[0060] Specifically, the first reset circuit unit generates a reset signal based on the reset command, and restarts the first processor and the second processor based on the reset signal.

[0061] In addition, the second reset circuit unit can also generate a reset signal based on the reset command, and send the reset signal to the first reset circuit unit through the connection backplane; the first reset circuit unit then restarts the first processor and the second processor based on the reset signal.

[0062] Step S650: the front-end module controls the first processor and the second processor to be in a standby state after restart.

[0063] Step S660: the first processor re-establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the third processor.

[0064] Step S670: The first processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0065] Specifically, when the first processor and the second processor are successfully restarted, the first processor and the second processor can provide redundant backup for the working third processor, thereby further improving the operation reliability of the server.

[0066] As attached Figure 3 As shown, the first reset circuit unit is the reset circuit in the front module; the second reset circuit unit is the reset circuit in the rear module; the first processor is HOST1, the second processor is HOST2, the third processor is HOST3, and the fourth processor is HOST4; the PCIE SWITCH in the front module is the first data exchange unit; the PCIE SWITCH in the rear module is the second data exchange unit; EP is a service device; EP and PCIE SWITCH, PCIE SWITCH and HOST1, PCIE SWITCH and HOST2, PCIE SWITCH and HOST3, and PCIE SWITCH and HOST4 are all connected through PCIE5.0X8 (8-channel PCIE5.0 bus) communication.

[0067] In a seventh embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-controller systems proposed by the present invention, based on the second embodiment, step S260 further includes the following steps: Step S710: When the third processor fails to operate, the fourth processor establishes a communication connection with the first device based on the NTB channel, and disconnects the communication connection between the first device and the third processor.

[0068] Step S720: The fourth processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0069] Specifically, when the third processor fails to operate, the fourth processor establishes a communication connection with the first device through the NTB channel (that is, the fourth processor immediately takes over the first device), and disconnects the communication connection between the first device and the third processor, and uses the fourth processor to seamlessly connect the computing tasks of the first device, thereby further improving the operating reliability of the server.

[0070] In an eighth embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-control proposed by the present invention, based on the second embodiment, step S210 further includes the following steps: Step S810: The third processor establishes a communication connection with the service device through the second data exchange unit to obtain service data sent by the service device, marks the service device communicatively connected to the third processor as a second device, and calculates the service data sent by the second device to implement the service function.

[0071] Step S820: When the third processor fails to operate, the fourth processor establishes a communication connection with the second device through the second data exchange unit, and disconnects the communication connection between the second device and the third processor.

[0072] Step S830: The fourth processor obtains the service data sent by the second device, and calculates the service data to implement the service function.

[0073] Step S840: the fourth processor sends a reset command to the second reset circuit unit.

[0074] Step S850: the second reset circuit unit restarts the third processor based on the reset command.

[0075] Specifically, the second reset circuit unit generates a reset signal based on the reset command, and restarts the third processor based on the reset signal.

[0076] Step S860: the rear module controls the third processor to be in a standby state after restart.

[0077] Specifically, through the scheme of this embodiment, daily computing tasks can be performed on the second device through the rear-inserted module to share the computing pressure of the front-inserted module; the third processor and the fourth processor back up each other. When the third processor fails to operate, the fourth processor establishes a communication connection with the second device through the second data exchange unit (that is, the fourth processor immediately takes over the second device), and disconnects the communication connection between the third device and the second processor, so that the fourth processor seamlessly connects the computing tasks for the second device.

[0078] In a ninth embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-controller systems proposed by the present invention, based on the first embodiment, step S120 further includes the following steps: Step S910: The front-end module obtains operating status parameters of the first processor, where the operating status parameters include occupancy, operating frequency, and operating temperature.

[0079] Step S920: The front-end module obtains the average occupancy rate, average operating frequency, and average operating temperature of the first processor within a preset period of time in the past based on the operating status parameters.

[0080] Step S930: When the operating status parameters of the first processor meet the fourth preset condition, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor, wherein the fourth preset condition is: the average occupancy rate within the past preset time period is greater than the first preset occupancy rate (for example, 90%), the average operating frequency within the past preset time period is greater than the preset frequency (the preset frequency in this embodiment is set to 90% of the rated frequency of the first processor), and the average operating temperature within the past preset time period is greater than the preset temperature (for example, 85°C).

[0081] Step S940: The second processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

[0082] Specifically, when the first processor meets the fourth preset condition, it means that the first processor is in a high-load operation state. In order to ensure the life of the first processor and prevent the first processor from being abnormally damaged, the first device is directly taken over by the second processor, and the first processor is put into standby state, thereby balancing the workload of the first processor and the second processor.

[0083] In a tenth embodiment of a high-reliability redundant backup calculation method based on multi-active and multi-controller proposed by the present invention, based on the first embodiment, step S120 further includes the following steps: Step S1010: the first processor marks a service device that has established a communication connection with the first processor and has not sent service data as a third device.

[0084] Specifically, the third device here is the object device on which the first processor is about to perform the computing task.

[0085] Step S1020: The first processor obtains the size of the business data that needs to be calculated by the third device and the required calculation time limit.

[0086] Step S1030: The first processor calculates the estimated occupancy rate of the first processor when calculating the business data of the third device based on the size of the business data to be calculated and the required calculation time limit.

[0087] Specifically, based on the size of the business data to be calculated and the required calculation time limit, the estimated occupancy rate of the first processor when calculating the business data of the third device can be obtained. The larger the size of the business data, the higher the corresponding estimated occupancy rate, and the shorter the required calculation time limit, the higher the corresponding estimated occupancy rate.

[0088] Step S1040: When the estimated occupancy rate is greater than a second preset occupancy rate (eg, 98%), the second processor establishes a communication connection with the third device through the first data exchange unit, and disconnects the communication connection between the third device and the first processor.

[0089] Step S1050: the second processor obtains the service data sent by the third device, and calculates the service data to implement the service function.

[0090] Specifically, when the estimated occupancy rate is greater than the second preset occupancy rate, it means that the first processor will have a hard time executing the computing tasks of the third device subsequently. In order to ensure the life of the first processor and prevent abnormal damage to the first processor, the third device is directly taken over by the second processor (that is, the computing tasks corresponding to the third device are executed by the second processor), thereby balancing the workload of the first processor and the second processor.

[0091] The present invention also proposes a high-reliability redundant backup computing system based on multi-active and multi-control, which applies a high-reliability redundant backup computing method based on multi-active and multi-control; the system includes a front-end module; the front-end module includes a first processor, a second processor and a first data exchange unit; the first processor and the second processor are both communicatively connected to the first data exchange unit; the first data exchange unit is communicatively connected to a business device.

[0092] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0093] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A high-reliability redundant backup calculation method based on multi-active and multi-control, characterized in that: Applicable to a high-reliability redundant backup computing system based on multi-active and multi-control; the system includes a front-end module; the front-end module includes a first processor, a second processor and a first data exchange unit; The first processor and the second processor are both communicatively connected to the first data exchange unit; the first data exchange unit is communicatively connected to the service device; the method includes: When the front plug-in module is started and the first preset condition is met, the front plug-in module controls the first processor to be in a running state and controls the second processor to be in a standby state, wherein the first preset condition is that both the first processor and the second processor can operate normally; The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, marks the service device communicatively connected to the first processor as the first device, and calculates the service data sent by the first device to implement the service function; When the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor; The second processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

2. According to claim 1, a high-reliability redundant backup calculation method based on multi-active and multi-controller is characterized in that: The system further includes a rear-insertion module and a connection backplane; the front-insertion module further includes a first storage unit; the first processor and the second processor are both communicatively connected to the first storage unit; the rear-insertion module includes a third processor, a fourth processor, a second storage unit and a second data exchange unit; the third processor and the fourth processor are both communicatively connected to the second data exchange unit; the third processor and the fourth processor are both communicatively connected to the second storage unit; the second data exchange unit is communicatively connected to a service device; the front-insertion module and the rear-insertion module are both communicatively connected to the connection backplane; the method further includes: After the rear module is started, the third processor and the fourth processor are controlled to be in a standby state; The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, marks the service device connected to the first processor as a first device, and calculates the service data sent by the first device to implement the service function, and then further includes: The first processor calculates the business data to obtain result data and stores it in the first storage unit; The front-insertion module sends the result data in the first storage unit to the second storage unit for backup storage via the connection backplane; When both the first processor and the second processor have operation failures, the third processor establishes an NTB channel through the second data exchange unit, the connection backplane and the first data exchange unit; The third processor establishes a communication connection with the first device based on the NTB channel, and disconnects the communication connection between the first device and the first processor; The third processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

3. The high-reliability redundant backup calculation method based on multi-active and multi-control according to claim 1 is characterized in that: The front-end module further includes a first reset circuit unit; the first processor and the second processor are both communicatively connected to the first reset circuit unit; when the first processor fails to operate, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor, and then further includes: The second processor sends a reset command to the first reset circuit unit; The first reset circuit unit restarts the first processor based on the reset command; The front-end module controls the first processor to be in a standby state after restart.

4. A high-reliability redundant backup calculation method based on multi-active and multi-control according to claim 3, characterized in that: The front-end module controls the first processor to be in a standby state after restart, and then further includes: When the second processor fails to operate, the first processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the second processor; The first processor obtains the service data sent by the first device, and calculates the service data to implement the service function; The first processor sends a reset command to the first reset circuit unit; The first reset circuit unit restarts the second processor based on the reset command; The front-end module controls the second processor to be in a standby state after restart.

5. The high-reliability redundant backup calculation method based on multi-active and multi-control according to claim 1 is characterized in that: Also includes: When the front plug-in module is started and the second preset condition is met, the front plug-in module controls the first processor to be in a running state and marks the second processor as a faulty processor, wherein the second preset condition is: the first processor can operate normally, and the second processor cannot operate normally; The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, and calculates the service data to implement the service function; When the front plug-in module is started and the third preset condition is met, the front plug-in module controls the second processor to be in a running state and marks the first processor as a faulty processor, wherein the third preset condition is: the second processor can operate normally, and the first processor cannot operate normally; The second processor establishes a communication connection with the service device through the first data exchange unit to obtain the service data sent by the service device, and calculates the service data to implement the service function.

6. A high-reliability redundant backup calculation method based on multi-active and multi-control according to claim 2, characterized in that: The rear-insertion module further includes a second reset circuit unit; the third processor and the fourth processor are both communicatively connected to the second reset circuit unit; The third processor obtains the service data sent by the first device, and calculates the service data to implement the service function, and then further includes: The third processor sends a reset command to the second reset circuit unit; The second reset circuit unit sends the reset command to the first reset circuit unit via the connection backplane; The first reset circuit unit restarts the first processor and the second processor based on the reset command; The front-insertion module controls the first processor and the second processor to be in a standby state after restart; The first processor re-establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the third processor; The first processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

7. A high-reliability redundant backup calculation method based on multi-active and multi-control according to claim 2, characterized in that: The third processor obtains the service data sent by the first device, and calculates the service data to implement the service function, and then further includes: When the third processor fails to operate, the fourth processor establishes a communication connection with the first device based on the NTB channel, and disconnects the communication connection between the first device and the third processor; The fourth processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

8. The high-reliability redundant backup calculation method based on multi-active and multi-control according to claim 2 is characterized in that: After the rear plug-in module is started, the third processor and the fourth processor are controlled to be in a standby state, and then the method further includes: The third processor establishes a communication connection with the service device through the second data exchange unit to obtain service data sent by the service device, marks the service device connected to the third processor as a second device, and calculates the service data sent by the second device to implement the service function; When the third processor fails to operate, the fourth processor establishes a communication connection with the second device through the second data exchange unit, and disconnects the communication connection between the second device and the third processor; The fourth processor obtains the service data sent by the second device, and calculates the service data to implement the service function; The fourth processor sends a reset command to the second reset circuit unit; The second reset circuit unit restarts the third processor based on the reset command; The rear module controls the third processor to be in a standby state after the restart.

9. The high-reliability redundant backup calculation method based on multi-active and multi-control according to claim 1 is characterized in that: The first processor establishes a communication connection with the service device through the first data exchange unit to obtain service data sent by the service device, marks the service device connected to the first processor as a first device, and calculates the service data sent by the first device to implement the service function, and then further includes: The front-end module obtains operating status parameters of the first processor, wherein the operating status parameters include occupancy rate, operating frequency and operating temperature; The front-end module obtains the average occupancy rate, average operating frequency and average operating temperature of the first processor within a preset time period in the past based on the operating status parameters; When the operating state parameter of the first processor meets the fourth preset condition, the second processor establishes a communication connection with the first device through the first data exchange unit, and disconnects the communication connection between the first device and the first processor, wherein the fourth preset condition is: the average occupancy rate within the past preset time period is greater than the first preset occupancy rate, the average operating frequency within the past preset time period is greater than the preset frequency, and the average operating temperature within the past preset time period is greater than the preset temperature; The second processor obtains the service data sent by the first device, and calculates the service data to implement the service function.

10. A highly reliable redundant backup computing system based on multi-active and multi-controller, characterized in that: A high-reliability redundant backup computing method based on multi-active and multi-control as described in any one of claims 1-9; the system includes a front-end module; the front-end module includes a first processor, a second processor and a first data exchange unit; the first processor and the second processor are both communicatively connected to the first data exchange unit; the first data exchange unit is communicatively connected to a business device.

Images