High-reliability PCS communication gateway dual-computer hot backup and seamless switching method

Abstract

The application discloses a high-reliability PCS communication gateway double-machine hot backup and seamless switching method, and relates to the technical field of PCS communication gateway, solves the problem that a task execution time sequence cannot be reasonably distributed according to a history operation law of a host computer, and problems such as a total task execution time and a backup time being not matched, task loss, and execution sequence disorder, after the host computer is confirmed to be dead, the backup computer can directly inherit and copy the operation task of the host computer, and it is not necessary to reestablish a task connection; meanwhile, by analyzing a history backup time of the host computer, the copied task is evenly distributed, a scientific backup task execution logic column is generated, a standard task and a split non-standard task are reasonably combined, the total task execution time and the backup time are matched, and all the copied tasks can be orderly executed in standby time of the backup computer.

Description

Technical Field

[0001] This invention relates to the field of PCS communication gateway technology, specifically to a highly reliable method for dual-machine hot standby and seamless switching of PCS communication gateways. Background Technology

[0002] As the core communication hub in power storage and new energy grid connection systems, the PCS communication gateway undertakes key tasks such as protocol conversion, data acquisition, and control command issuance. Its operational stability directly determines the normal operation of the entire PCS system.

[0003] Currently, existing PCS communication gateway dual-machine hot standby switching technology still has many shortcomings and cannot meet the actual requirements of high reliability and seamless switching. In practical applications, when the primary machine fails and needs to switch to the standby machine, there is often a problem of poor task handover: the standby machine cannot directly inherit the operation tasks of the primary machine and needs to re-establish task connections, resulting in service interruption during the switching process; Meanwhile, the lack of a scientific task allocation mechanism means that after the standby machine replicates the host machine's tasks, it cannot reasonably allocate the task execution sequence according to the host machine's historical operating patterns. This can easily lead to situations such as a mismatch between the total task execution time and the standby machine's cache time, task loss, and disordered execution order, which in turn can cause problems such as message loss, failure of control command issuance, and interruption of data acquisition.

[0004] In addition, in some switchover schemes, the backup machine's replication of the master machine's tasks lacks specificity. It continues to replicate tasks regardless of whether the master machine has actually crashed. This not only wastes backup machine resources but may also cause system instability due to blind switching, further affecting the operational reliability of the PCS communication gateway.

[0005] To address the shortcomings of the existing technologies, there is an urgent need for a dual-machine hot standby and seamless switching method that can achieve seamless task connection, reasonable task allocation, and accurate identification of host status. This method would solve problems such as switching interruption, task loss, and resource waste in the existing technologies and ensure the stable operation of the PCS system. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a highly reliable method for dual-machine hot standby and seamless switching of PCS communication gateways. This method solves the problems of being unable to reasonably allocate task execution time according to the host's historical operating patterns, which can easily lead to mismatches between the total task execution time and the standby machine's cache time, task loss, and disordered execution order.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a highly reliable method for dual-machine hot standby and seamless switching of a PCS communication gateway, comprising the following steps: Step 1: Monitor the operating parameters of the host corresponding to the PCS communication gateway in real time, and based on the real-time monitored operating parameters, assess whether the host is in an abnormal state. The specific method is as follows: The operating parameters of the corresponding host are monitored in real time, and the real-time monitored operating parameters are compared with the monitoring range set for the corresponding monitoring item. If all the real-time monitored operating parameters are within the set monitoring range, no processing is required, and monitoring can continue. If there are any monitoring items that are not within the monitoring range, an abnormal signal is recorded. The abnormal time T of the abnormal signal is confirmed in real time. If T ≥ 3s, the host is marked as abnormal; otherwise, monitoring continues. Step 2: When the host machine is in an abnormal state, confirm the heartbeat characteristics generated by the host machine and simultaneously control the standby machine to replicate the host machine's operation tasks in real time. Identify whether there are excessively long intervals in the confirmed heartbeat characteristics. If so, determine that the host machine is in a frozen state; if not, control the standby machine to delete the replicated host machine operation tasks. The specific method is as follows: When the host is in an abnormal state, the current time is recorded as the abnormal time. Based on the determined abnormal time, a set of tracing cycles is locked. The tracing cycle is a preset cycle. The interval duration associated between historical adjacent heartbeats is identified within the tracing cycle. The minimum value is selected from several sets of recorded interval durations and recorded as the standard delay. Based on the confirmed standard delay Ts, a set of evaluation ranges is locked, and the confirmed abnormal moment is recorded as Tq. The determined evaluation range is [Tq, Tq+3Ts]. Within the determined time evaluation range, it is confirmed whether the standby machine receives the heartbeat of the master machine. If so, no processing is required, and the standby machine is controlled to delete the replicated master operation tasks. If not, the master machine is determined to be in a dead state, and the replicated master operation tasks of the standby machine are retained simultaneously. Step 3: When the primary machine is in a crash state, execute the primary / standby machine switchover process. Based on the requester associated with the primary machine, directly control the standby machine to execute the corresponding operation tasks. Then, based on the cache time of the primary machine in the historical process, evenly distribute the primary machine operation tasks replicated by the standby machine, lock the optimal allocation process, and determine the subsequent cached task execution logic. The specific method is as follows: Based on the confirmed anomaly time and the marked tracing period, the execution task data associated with the host within the tracing period is confirmed, and the standby duration associated with the host is confirmed from the associated execution task data. From the associated standby time periods, the cache time is selected: the associated standby durations are sorted in ascending order of value to confirm the standby duration sequence, and associated duration segments are randomly selected from the confirmed standby duration sequence. The associated duration segments are continuous data segments within the standby duration sequence, and the density of the selected associated duration segments is confirmed. The duration range F of the associated duration segments and the total standby duration G are locked, and the density M of the corresponding associated duration segment is confirmed using: F÷G=M. Then, the density M associated with different associated duration segments within the standby duration sequence is confirmed in turn, and the minimum value is selected from the confirmed density M, and the associated duration segment associated with the minimum value is recorded as the determined duration segment. From the marked defined duration, the average of the standby durations present inside is processed, and the resulting average is recorded as the cache time. The specific method for evenly distributing the host operation tasks replicated by the backup machine is as follows: Confirm the required time associated with the host operation tasks replicated by the standby machine, and compare the required time with the confirmed cache time. Host operation tasks that meet the requirement time ≤ cache time are recorded as qualified tasks, and host operation tasks that do not meet the requirement time ≤ cache time are recorded as unqualified tasks. Sort the required time of the target tasks in ascending order to confirm the demand time sequence. Starting from the first demand in the demand time sequence, identify whether the total time of the first demand time and the subsequent demand times satisfy the following conditions: Total time > Cache time. If so, the target task associated with the first demand time is recorded as a single task. If the total time < Cache time, the target task associated with the first demand time and the subsequent demand times is recorded as a combined task. Continue to confirm the combined tasks until the total time > Cache time. Record the target tasks associated with several groups of demand times before the corresponding demand time as combined tasks. If the total time = Cache time, record the target tasks associated with two groups of demand times as combined tasks. The specific method for confirming the execution logic of cached tasks is as follows: Record the confirmed combined tasks as tasks to be executed, and associate the tasks to be executed with a single cache execution signal. Then, split the unqualified tasks into tasks, combine the split tasks with the single tasks, lock the combined tasks, and the total time associated with the combined tasks must meet the following condition: total time ≤ cache time. Synchronously record the confirmed combined tasks as tasks to be executed, and synchronously associate them with a single cache execution signal. The associated split tasks have the task tag associated with the original unqualified tasks. The remaining split tasks of the unqualified tasks are then recorded as tasks to be executed, and the associated single-cache execution signal is synchronized. The confirmed tasks to be executed are collected and sorted according to time order to generate a cached task execution logic column; Step 4: For the determined cache task execution logic, during the standby time associated with the standby machine, execute the corresponding cache task operation process. The specific method is as follows: The cached tasks are executed in the execution logic column. The corresponding tasks are executed during the standby time of the standby machine. The system identifies whether there is a single cached execution signal for the task. If there is, the cached execution is performed. If not, the task is determined to be a split task and the associated tasks are determined, which are combined tasks. The tasks are executed during the associated single standby time. Each set of standby time executes a set of tasks. In step four, after the task to be executed completes the transmission process, the unqualified tasks are combined according to the task tags associated with the split tasks and the unqualified tasks to generate the original task data. The task tags include the front and back tags.

[0008] This invention provides a highly reliable method for dual-machine hot standby and seamless switching of PCS communication gateways. Compared with existing technologies, it has the following advantages: After the primary host is determined to be in an abnormal state, a switchover is not initiated immediately. Instead, a standard delay is determined by tracing historical heartbeat data, and a reasonable evaluation range is set based on the time of the abnormality to determine whether the heartbeat interval of the primary host is too long, thereby accurately determining whether the primary host is in a dead state. This method avoids the waste of resources caused by the backup machine blindly intervening when the primary host is not truly dead (such as a brief failure or a momentary link interruption), and also prevents service interruptions caused by the failure to identify the primary host dead in a timely manner and the delay in switchover. At the same time, the backup machine only retains the replication task after confirming that the primary host is dead, and deletes the replication task when the primary host is not dead, which effectively reduces the resource consumption of the backup machine and improves the system operating efficiency. After confirming a primary system crash, the standby system can directly inherit the replicated primary system operation tasks without re-establishing task connections. Simultaneously, by analyzing the primary system's historical cache time, replicated tasks are evenly distributed, generating a scientifically structured cache task execution logic. This rationally combines compliant tasks and split non-compliant tasks, ensuring the total task execution time matches the standby system's cache time. This allows the standby system to execute all replicated tasks in an orderly manner during standby time. This design effectively avoids issues such as packet loss, task omissions, and disordered execution order during the switchover process, ensuring uninterrupted communication between the PCS communication gateway and the requester, and smooth data transmission. It also guarantees the normal issuance of PCS system scheduling and control commands and the continuity of data acquisition. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0010] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0011] First Embodiment Please see Figure 1 This application provides a highly reliable method for dual-machine hot standby and seamless switching of a PCS communication gateway, including the following steps: Step 1: Monitor the operating parameters of the host corresponding to the PCS communication gateway in real time, and assess whether the host is in an abnormal state based on the real-time monitoring operating parameters. Specifically, when the corresponding host has abnormal parameters, it means that the corresponding host is in an abnormal state. Each different monitoring parameter has a corresponding monitoring benchmark. According to the corresponding monitoring benchmark, it is possible to effectively confirm whether the corresponding operating parameter is abnormal, thereby assessing whether the corresponding host is abnormal. Step 2: When the host is in an abnormal state, the heartbeat characteristics generated by the host are confirmed, and the standby machine is synchronously controlled to replicate the host operation tasks in real time. The heartbeat characteristics are identified to see if there are excessively long intervals. If so, the host is determined to be in a dead state. If not, the standby machine is controlled to delete the replicated host operation tasks. Step 3: When the host is in a crash state, execute the switchover process between the host and the standby host, and directly control the standby host to execute the corresponding operation tasks based on the requester associated with the host. Then, based on the cache time of the host in the historical process, distribute the host operation tasks replicated by the standby host evenly, lock the optimal allocation process, and determine the subsequent cache task execution logic.

[0012] Second Embodiment In the specific implementation process, compared with the above embodiments, this embodiment mainly focuses on the evaluation process of abnormal host status; The specific methods for conducting the evaluation are as follows: The operating parameters of the corresponding host are monitored in real time, and the real-time monitored operating parameters are compared with the monitoring range set for the corresponding monitoring item. If all the real-time monitored operating parameters are within the set monitoring range, no processing is required, and monitoring can continue. If there are any monitoring items that are not within the monitoring range, an abnormal signal is recorded. The abnormal time T of the abnormal signal is confirmed in real time. If T ≥ 3s, the host is marked as abnormal; otherwise, monitoring continues. Specifically, during the data monitoring and processing process, there may be data fluctuations. These fluctuations will cause corresponding abnormal signals. However, the duration of these abnormal signals is usually short, so there is no need to label the abnormal state. But if the duration of the abnormal signal is too long, it means that the corresponding host is in an abnormal state, and subsequent analysis processes need to be performed to determine whether a primary / backup switchover process is necessary.

[0013] Third Embodiment In this embodiment, compared to the above embodiments, this embodiment mainly focuses on the specific identification process in step two. The specific method for determining whether there are excessively long intervals in heartbeat characteristics is as follows: When the host is in an abnormal state, the current time is recorded as the abnormal time. Based on the determined abnormal time, a set of tracing cycles is locked. The tracing cycle is a preset cycle, and its specific value is determined by the operator based on experience. Generally, it is 24 hours. The interval duration associated between historical adjacent heartbeats (corresponding transmission signals with timestamps) is identified within the tracing cycle. The minimum value is selected from several sets of recorded interval durations and recorded as the standard delay. Based on the confirmed standard delay Ts, a set of evaluation ranges is locked, and the confirmed abnormal moment is recorded as Tq. The determined evaluation range is [Tq, Tq+3Ts]. Within the determined time evaluation range, it is confirmed whether the standby machine receives the heartbeat of the master machine. If so, no processing is required, and the standby machine is controlled to delete the replicated master operation tasks. If not, the master machine is determined to be in a dead state, and a switchover signal is generated, while the replicated master operation tasks of the standby machine are retained simultaneously. The host operation tasks generally include the execution tasks that the corresponding host participates in, such as sending messages, related network configurations, etc. As long as the host has operation tasks from other requesters, its standby machine will execute a set of replication processes, but will not send them, only record them. If it is subsequently confirmed that the corresponding host is in a dead state, then the standby machine will intervene. In the subsequent intervention process, it will execute the execution tasks associated with the corresponding host. However, during the heartbeat interval, it is not entirely certain whether there are any unexecuted operations on the corresponding host. Therefore, the standby machine replicates the corresponding operation tasks and executes the replicated operation tasks and performs related transmissions during the buffer time in the subsequent operation process.

[0014] Fourth embodiment In this embodiment, compared to the above embodiments, this embodiment mainly focuses on the specific processing flow in step three. The specific method for confirming the cache time of the host in historical processes is as follows: Based on the confirmed anomaly time and the marked tracing period, the execution task data associated with the host within the tracing period is confirmed, and the standby duration associated with the host is confirmed from the associated execution task data. From the associated standby time periods, the cache time is selected: the associated standby durations are sorted in ascending order of value to confirm the standby duration sequence, and associated duration segments are randomly selected from the confirmed standby duration sequence. The associated duration segments are continuous data segments within the standby duration sequence, and the density of the selected associated duration segments is confirmed. The duration range F of the associated duration segments and the total standby duration G are locked, and the density M of the corresponding associated duration segment is confirmed using: F÷G=M. Then, the density M associated with different associated duration segments within the standby duration sequence is confirmed in turn, and the minimum value is selected from the confirmed density M, and the associated duration segment associated with the minimum value is recorded as the determined duration segment. From the marked and defined time periods, the average of the standby time within the host is processed, and the resulting average is recorded as the cache time. Specifically, the cache time is the rest time associated with the corresponding host executing the relevant scheduled task of the corresponding requester. During the corresponding rest time, the corresponding clustering logic is executed, the optimal solution is locked from the clustering processing features, and the corresponding cache time is locked from the locked optimal solution. This facilitates the effective execution of the host operation tasks existing on the standby machine, ensuring that the corresponding host operation tasks can be effectively delivered to the corresponding requester. The specific method for confirming the execution logic column of the cache task is as follows: Confirm the required time associated with the host operation tasks replicated by the standby machine, and compare the required time with the confirmed cache time. Host operation tasks that meet the requirement time ≤ cache time are recorded as qualified tasks, and host operation tasks that do not meet the requirement time ≤ cache time are recorded as unqualified tasks. Sort the required time of the target tasks in ascending order to confirm the demand time sequence. Starting from the first demand in the demand time sequence, identify whether the total time of the first demand time and the subsequent demand times satisfy the following conditions: Total time > Cache time. If so, the target task associated with the first demand time is recorded as a single task. If the total time < Cache time, the target task associated with the first demand time and the subsequent demand times is recorded as a combined task. Continue to confirm the combined tasks until the total time > Cache time. Record the target tasks associated with several groups of demand times before the corresponding demand time as combined tasks. If the total time = Cache time, record the target tasks associated with two groups of demand times as combined tasks. Record the confirmed combined tasks as tasks to be executed, and associate the tasks to be executed with a single cache execution signal. Then, split the unqualified tasks into tasks, combine the split tasks with the single tasks, lock the combined tasks, and the total time associated with the combined tasks must meet the following condition: total time ≤ cache time. Synchronously record the confirmed combined tasks as tasks to be executed, and synchronously associate them with a single cache execution signal. The associated split tasks have the task tag associated with the original unqualified tasks. The remaining split tasks of the unqualified tasks are then recorded as tasks to be executed, and the associated single-cache execution signal is synchronized. The confirmed tasks to be executed are collected and processed in sequence, and sorted according to time to generate a cached task execution logic column.

[0015] Fifth Embodiment In this embodiment, compared to the above embodiments, the main focus is on the task execution process of the standby machine. Step 4: For the determined cache task execution logic column, execute the corresponding cache task operation process within the standby time associated with the standby machine; For tasks to be executed within the cached task execution logic column, cache execution processing is performed. During the standby time of the standby machine, the corresponding tasks to be executed are executed. It is determined whether there is a single cache execution signal for the tasks to be executed. If there is, cache execution processing is performed. If not, the task is determined to be a split task, and the associated tasks to be executed are determined, which are combined tasks (in a few cases, combined tasks may be lost). The tasks to be executed are executed within the associated single standby time. Each set of standby time executes one set of tasks to be executed. Subsequently, based on the task tags associated with the split tasks and the unsatisfactory tasks, the unsatisfactory tasks are combined to generate the original message and other task data. The task tags include the front and back tags, because the splitting involves a front and back sorting.

[0016] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0017] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A highly reliable method for dual-machine hot standby and seamless switching of a PCS communication gateway, characterized in that, Includes the following steps: Step 1: Monitor the operating parameters of the PCS communication gateway host in real time, and determine whether the host has entered an abnormal state based on the duration of abnormal parameters. Step 2: When the primary machine malfunctions, the backup machine replicates the primary machine's operation tasks in real time and dynamically determines whether the primary machine has crashed based on the primary machine's historical heartbeat interval characteristics; if the primary machine has not crashed, the backup machine clears the replicated tasks. If the system crashes, retain the replicated task and initiate a primary / standby switchover. Step 3: After the primary / standby switchover, extract the cache time characteristics of the primary machine's historical operation, and perform adaptive allocation and reorganization of the primary machine's replicated tasks on the standby machine to generate a task execution sequence that matches the cache time. Step 4: The standby machine executes all replication tasks and restores data in an orderly manner within its own standby window according to the generated task execution sequence, achieving seamless switching.

2. The highly reliable PCS communication gateway dual-machine hot standby and seamless switching method according to claim 1, characterized in that, In step one, the specific method for assessing whether the host is in an abnormal state is as follows: The system monitors the host's operating parameters in real time and compares them with preset threshold ranges. When a parameter exceeds the threshold, an abnormal signal is recorded. When the duration of the abnormal signal reaches a preset time threshold, the host is marked as being in an abnormal state.

3. The highly reliable PCS communication gateway dual-machine hot standby and seamless switching method according to claim 1, characterized in that, In step two, the specific method for dynamically determining whether the host has crashed based on the host's historical heartbeat interval characteristics is as follows: The current moment when the abnormal state is recorded as the abnormal moment, and based on the determined abnormal moment, a set of tracing cycles is locked. The tracing cycle is a preset cycle. The interval duration associated between historical adjacent heartbeats is identified from the tracing cycle, and the minimum value is selected and recorded as the standard delay. Based on the standard delay Ts, a set of evaluation ranges is locked, and the confirmed abnormal moment is denoted as Tq. The determined evaluation range is [Tq, Tq+3Ts]. Within the determined time evaluation range, it is confirmed whether the standby machine receives the heartbeat of the master machine. If so, no processing is required, and the standby machine is controlled to delete the replicated master operation tasks. If not, the master machine is determined to be in a dead state, and the replicated master operation tasks of the standby machine are retained simultaneously.

4. The highly reliable PCS communication gateway dual-machine hot standby and seamless switching method according to claim 1, characterized in that, In step three, the specific method for extracting cache time features is as follows: Based on the confirmed anomaly time and the marked tracing period, the execution task data associated with the host within the tracing period is confirmed, and the standby duration associated with the host is confirmed. From several sets of standby time periods, the cache time is selected: the several sets of standby durations are sorted in ascending order of value to confirm the standby duration sequence, and associated duration segments are randomly selected from them. The associated duration segments are continuous data segments within the standby duration sequence. The density of the selected associated duration segments is confirmed, the duration range F of the associated duration segments and the total standby duration G are locked, and the density M of the corresponding associated duration segment is confirmed by using: F÷G=M. Then, the density M associated with different associated duration segments within the standby duration sequence is confirmed in turn, and the minimum value is selected from several confirmed density M groups. The associated duration segment associated with the minimum value is recorded as the determined duration segment. From the marked defined duration, the average of the standby durations is calculated, and the resulting average is recorded as the cache time.

5. The highly reliable PCS communication gateway dual-machine hot standby and seamless switching method according to claim 4, characterized in that, In step three, the specific method for evenly distributing the host operation tasks replicated by the backup machine is as follows: Confirm the required time associated with the host operation tasks replicated by the standby machine, and compare the required time with the confirmed cache time. Host operation tasks that meet the requirement time ≤ cache time are recorded as qualified tasks, and host operation tasks that do not meet the requirement time ≤ cache time are recorded as unqualified tasks. Sort the required time of the target tasks in ascending order to confirm the demand time sequence. Starting from the first demand in the demand time sequence, identify whether the total time of the first demand time and the subsequent demand times satisfy the following conditions: Total time > Cache time. If so, the target task associated with the first demand time is recorded as a single task. If the total time < Cache time, the target task associated with the first demand time and the subsequent demand times is recorded as a combined task, and the confirmation of combined tasks is carried out until the total time > Cache time. The target tasks associated with several groups of demand times before the corresponding demand time are recorded as combined tasks. If the total time = Cache time, the target tasks associated with two groups of demand times are recorded as combined tasks.

6. The highly reliable PCS communication gateway dual-machine hot standby and seamless switching method according to claim 5, characterized in that, In step three, the specific method for confirming the execution logic of the cache task is as follows: Record the confirmed combined tasks as tasks to be executed, and associate the tasks to be executed with a single cache execution signal. Then, split the unqualified tasks into tasks, combine the split tasks with the single tasks, lock the combined tasks, and the total time associated with the combined tasks must meet the following condition: total time ≤ cache time. Synchronously record the confirmed combined tasks as tasks to be executed, and synchronously associate them with a single cache execution signal. The associated split tasks have the task tag associated with the original unqualified tasks. The remaining split tasks of the unqualified tasks are then recorded as tasks to be executed, and the associated single-cache execution signal is synchronized. The confirmed tasks to be executed are collected and processed in sequence, and sorted according to time to generate a cached task execution logic column.

7. The highly reliable PCS communication gateway dual-machine hot standby and seamless switching method according to claim 1, characterized in that, In step four, the specific method for executing the corresponding cache task operation process is as follows: The pending tasks within the cached task execution logic column are cached and executed. During the standby time of the standby machine, the corresponding pending tasks are executed. It is determined whether there is a single cached execution signal for the pending task. If there is, cached execution is performed. If not, the task is determined to be a split task, and the associated pending tasks, i.e., combined tasks, are determined. The pending tasks are executed within the associated single standby time. Each set of standby time executes a set of pending tasks.

8. The highly reliable PCS communication gateway dual-machine hot standby and seamless switching method according to claim 7, characterized in that, In step four, after the task is completed, the split subtasks are reassembled into the original complete task based on the preceding and following association markers of the split task, thus restoring the original data structure of the task.

Images