Vehicle-mounted cloud platform application high-availability system and deployment method

Abstract

The invention provides a vehicle-mounted cloud platform application high-availability system and a deployment method. The system comprises a cloud application unit, a control unit, an execution unit, a storage unit and a data unit, the overall high availability of the cloud platform system is adjusted in time through state monitoring and feedback of all the units and the load condition, and the vehicle-mounted cloud platform system and vehicle-mounted application are covered. Through the high-availability design of the application unit, the control unit, the execution unit, the storage unit and the data unit, automatic maintenance of the vehicle-mounted application of the cloud platform in the application online stage is realized, and high availability and high reliability of the vehicle-mounted application are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-mounted cloud platform applications, and in particular to a vehicle-mounted cloud platform application high-availability system and a deployment method. Background Art

[0002] In the national smart transportation, rail transit and other industries, moving vehicle-mounted applications to the cloud can effectively solve the problem of intensive resource management in embedded environments, achieve coordinated resource scheduling of applications, high-quality, continuous and stable operation, and unified lifecycle management.

[0003] In actual production projects, the migration of embedded vehicle applications to the cloud is divided into two stages, mainly including the debugging and adaptation stage of the vehicle application on the cloud platform and the online operation stage of the vehicle application.

[0004] During the debugging and adaptation stage of vehicle-mounted applications, the main work is the installation and deployment of the cloud platform and the linkage debugging of vehicle-mounted applications and vehicle-mounted equipment. Manual operation is required. The work in this stage is mainly carried out in the integration test workshop. The power and network environment are stable, and the cloud platform environment is often relatively stable, which can meet the adaptation needs of vehicle-to-vehicle applications.

[0005] During the on-board application online operation stage, the on-board cloud platform and on-board applications are running on the train, and there is often no human operation and maintenance. The on-board applications need to run stably without supervision. In the on-board mobile operation environment, the cloud platform's cloud controller, node controller, cloud storage, cloud database and other key components need to be deployed in a primary and backup manner for redundancy and reliability considerations, and distributed in multiple carriages. When there are environmental anomalies such as unstable power supply and unstable network connection, the above-mentioned key components may fail. More seriously, after the abnormal environment such as power failure and network disconnection is restored, the above-mentioned components will have abnormalities such as brain split, causing the on-board cloud application to be unable to continue to provide services.

[0006] To ensure high availability of in-vehicle cloud platform applications, the industry currently focuses on designing and ensuring the availability of cloud platforms and in-vehicle applications:

[0007] In terms of the vehicle-mounted cloud platform, the container cloud high-availability architecture is designed in the master-slave mode. Its core design relies on the synchronization of metadata between the master and slave nodes. The master node provides the container cloud platform control plane service, and the slave node synchronizes and backs up the container cloud platform metadata in real time. When the master node is unavailable, the master-slave switching model of the container cloud platform is used to quickly switch the cloud platform control plane components to the slave node, such as Figure 1 shown.

[0008] In terms of in-vehicle applications, they are mainly based on container instances. Applications are first deployed in container images and stored in image repositories. The node controller monitors the status of the operating environment of the in-vehicle application and reports its status to the cloud controller. When a container instance is abnormal, a new container instance is rescheduled to complete the startup of the application, thereby ensuring the sustainability of the application service.

[0009] In the in-vehicle embedded environment, whether in the in-vehicle cloud platform or in-vehicle applications, it is difficult to avoid the risk of brain split caused by network disconnection. The standby controller determines the status of the master controller through an independent heartbeat connection line. When the heartbeat connection between the master and standby controllers is interrupted due to a network failure and the state of each other cannot be determined, a brain split will occur, including brain splits in the cloud controller, cloud database, and cloud storage. As a result, the master and standby controllers will schedule the node controller to create a new instance, causing a catastrophic application dual-run failure, and the application cannot provide services normally. Summary of the invention

[0010] In view of the above situation, the present invention aims at the deficiencies of the prior art and provides a high-availability system and deployment method for an in-vehicle cloud platform application, aiming to be able to timely adjust the overall high availability of the cloud platform system, covering the in-vehicle cloud platform system and in-vehicle applications.

[0011] The present invention is achieved through the following technical solutions:

[0012] The present invention firstly provides a high-availability system for vehicle-mounted cloud platform applications, which includes: an application unit, a control unit, an execution unit, a storage unit and a data unit.

[0013] The control unit includes a cloud controller, which is distributed in two adjacent carriages in a master-slave deployment mode, and the cloud controller is used to send resource requests, receive application creation requests and schedule node controllers to process the creation of the application;

[0014] The application unit includes an application self-maintenance component, and the application self-maintenance component is used to accept a creation request of an application template issued by the cloud controller;

[0015] The storage unit includes cloud storage and a cloud storage monitoring module. The cloud storage is distributed one-to-one according to the cloud controller in a primary-backup manner and is used to provide storage volume mounting for the application unit; the cloud storage monitoring module is used to synchronize the primary and backup cloud storage and monitor the status of the cloud storage;

[0016] The data unit includes a cloud database, a cloud database backup module, and a cloud database monitoring module. The primary and backup cloud databases are distributed one-to-one according to the cloud controller in a primary-backup manner, and are mainly used to store resource data of the vehicle-mounted cloud platform; the cloud database backup module is used to back up the database resources of the vehicle-mounted cloud platform after the application adaptation phase is completed, and is saved as backup_init.tar.gz; the cloud database monitoring module is used to synchronize the primary and backup database data in real time and monitor the database status;

[0017] The execution unit includes a plurality of node controllers, which are distributed in each carriage. The node controllers are mainly responsible for processing the creation request sent by the cloud controller and saving it to the cloud database and the cloud storage.

[0018] According to the high-availability system for vehicle-mounted cloud platform applications, the application template saves the application information required by the vehicle-mounted cloud into the database during the application adaptation phase, including the images, storage volumes, and IP resources required by the application; during the application online phase, the application template automatically creates and monitors the application, and after an application failure, the application template actively sends an application creation request to the cloud controller.

[0019] According to the high-availability system applied on the vehicle cloud platform, the cloud controller realizes the master-slave capability through the distributed mechanism of Erlang; when the master controller is running, the standby controller stops; when the controlled device cannot connect to the master controller through the network heartbeat, the standby controller runs, thereby ensuring the master-slave high availability of the cloud controller.

[0020] According to the vehicle-mounted cloud platform application high-availability system, the cloud storage monitoring module is used to monitor the status of the cloud storage in real time, and automatically synchronize the primary and backup cloud storage data when an abnormality occurs in the primary and backup cloud storage to ensure the availability of the cloud storage.

[0021] According to the vehicle-mounted cloud platform application high-availability system, the cloud database monitoring module is used to monitor the status of the cloud database in real time, and when a brain split exception occurs in the primary and standby cloud controllers, the cloud database backup module backs up the file backup_init.tar.gz to restore data to ensure the availability of the cloud database.

[0022] The present invention also provides a deployment method for the vehicle-mounted cloud platform application high-availability system, and the specific steps of the deployment method are as follows:

[0023] After completing the cloud platform deployment on the embedded board and completing the linkage work between the vehicle-mounted application and the vehicle-mounted device in the adaptation phase, physical resources are created on the node controller and data is saved to the cloud storage and cloud database;

[0024] Sending a request to create the application template to the application self-maintenance component through the cloud controller;

[0025] The application self-maintenance component binds and creates the application template according to the logical resource information and saves it in the cloud database;

[0026] After the application template is successfully created, the cloud database backup operation is performed to back up the file backup_init.tar.gz, and the vehicle-mounted cloud platform is ready to enter the online operation stage;

[0027] The application template notifies the cloud controller to create an application. The cloud controller processes the creation of the application by scheduling it to the node controller, and saves the application resource information in the cloud database, and enters the online operation stage.

[0028] According to the deployment method of the vehicle-mounted cloud platform application high-availability system, when the application is down due to an abnormality, the application template notifies the cloud controller to reschedule to a new node controller to create the application;

[0029] When an abnormality occurs in the cloud controller, after the network is restored, the vehicle-mounted cloud platform restores the active and standby operation status of the cloud controller according to the node priority configured in advance, and the node with the highest priority becomes the active cloud controller;

[0030] When an exception occurs in the cloud storage, after the exception is recovered, the cloud storage monitoring module selects the primary storage through the virtual IP management component, and then synchronizes the data of the primary storage to the backup storage to restore the cloud storage state;

[0031] When the cloud data is abnormal, after the abnormality is recovered, the cloud database monitoring module restores the cloud database state from the backup file backup_init.tar.gz data.

[0032] According to the deployment method of the high-availability system for the in-vehicle cloud platform application, the physical resources created by the node controller include images, storage volumes, and network data; the application self-maintenance component binds and creates the application template based on logical resource information, and the logical resource information includes images, storage volumes, and network IP addresses.

[0033] According to the deployment method of the high-availability system for the in-vehicle cloud platform application, the abnormal application downtime includes downtime caused by abnormalities in the storage module, network or node controller; the abnormality in the cloud controller includes a network partition abnormality that causes a brain split abnormality in the cloud controller; the abnormality in the cloud storage includes frequent network dropouts and network disconnection abnormalities that cause a brain split abnormality in the cloud storage.

[0034] By means of the above technical solution, compared with the prior art, the beneficial effects of the present invention are:

[0035] 1) Through the high availability design of the above-mentioned application unit, control unit, execution unit, storage unit and data unit, the cloud platform realizes automatic maintenance of vehicle applications during the application launch phase, ensuring the high availability and high reliability of vehicle applications.

[0036] The debugging phase and the online operation phase of the cloud platform vehicle application are separated by design, so as to ensure the purity of the cloud platform environment and the certainty of resources during the online operation phase.

[0037] 2) Through the distributed, high-concurrency, and lightweight features of Erlang at the language level, lightweight and simple deployment of cloud controllers and node controllers can be achieved, while ensuring high availability of cloud controllers and node controllers.

[0038] 3) During the online operation phase of the cloud platform, application templates managed by application template components are used to achieve one-to-one creation and monitoring of applications, which can ensure high availability of vehicle-mounted applications under the premise of a normal cloud platform environment.

[0039] 4) When abnormalities such as power outages and network disconnections cause brain splits in key components of the cloud platform, the self-recovery capabilities of each component ensure the high reliability of the cloud platform, thereby ensuring the high availability of in-vehicle applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The diagram shows the high availability design of the active-standby mode of the existing vehicle-mounted cloud platform;

[0041] Figure 2 Shown is a structural block diagram of a high-availability system for a vehicle-mounted cloud platform application of the present invention;

[0042] Figure 3 Shown is a schematic diagram of a deployment method of a high-availability system for a vehicle-mounted cloud platform application according to the present invention;

[0043] Figure 4 Shown is a schematic diagram of high availability recovery of each module of the high availability system of the vehicle-mounted cloud platform application of the present invention. DETAILED DESCRIPTION

[0044] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] The technical concept of the present invention is: during the application adaptation and debugging stage, unpredictable operations such as frequent application creation, image packaging, and application configuration will be performed, which will have a certain impact on the stability of the vehicle-mounted cloud platform; however, in a production environment, it is only necessary to ensure the smooth operation of the application, without manual intervention, and all resources of the cloud platform remain determined.

[0046] Therefore, the solution of the present invention separates the application debugging phase and the online operation phase of the vehicle-mounted cloud platform system by design, and proposes a maintenance mode and a production mode for the vehicle-mounted cloud platform, thereby effectively solving the problem of misaligned requirements for cloud platform capabilities in the debugging phase and the production phase.

[0047] In maintenance mode, it is mainly responsible for application debugging, image packaging, testing, etc., and prepares the images, storage, network, application templates and other resources required for the application in the production phase; in production mode, through the high-availability design of application units, control units, execution units, storage units and data units, the stable operation of the application is guaranteed without human intervention.

[0048] By running the above two modes separately, the vehicle-mounted cloud platform can automatically maintain the vehicle-mounted applications during the application launch phase, ensuring the high availability and reliability of the vehicle-mounted applications.

[0049] Reference Figure 1 As shown, an embodiment of the present invention provides a structural block diagram of a high-availability system for a vehicle-mounted cloud platform application, and a high-availability system for a vehicle-mounted cloud platform system includes:

[0050] The vehicle-mounted cloud platform 300 includes an application unit 301, a control unit 302, an execution unit 303, a data unit 304 and a storage unit 305; among them, the cloud controller, the cloud database in the master-slave mode, the cloud storage in the master-slave mode and the node controller in the distributed mode, each component supports stand-alone independent operation, master-slave switching and automatic recovery.

[0051] The control unit includes a cloud controller 307 with a master-slave mode, which is distributed in two adjacent carriages through a master-slave deployment method. The cloud controller is used to send resource requests, receive requests from application units to create applications, and schedule node controllers to process the application component requests to create container applications. The master and standby cloud controllers are deployed in two adjacent carriages, respectively, and the master-standby capabilities of the cloud controller are realized through the distributed mechanism of Erlang. When the master controller is running, the standby controller stops; when the controlled device cannot connect to the master controller through the network heartbeat, the standby controller runs, thereby ensuring the master-standby high availability of the cloud controller.

[0052] The application unit includes an application self-maintenance component 306, which is used to accept the application template creation request issued by the cloud controller. The application template is created and saved in the database according to the application information required by the vehicle cloud during the application adaptation stage, including the image, storage volume, IP resources, etc. required by the application; during the application online stage, the application template automatically creates and monitors the application. When the application fails due to network, storage, etc., the application template component actively sends an application creation request to the cloud controller to create a new application according to the configured information.

[0053] The execution unit includes multiple node controllers 308, which are distributed in each car. The node controller is mainly responsible for processing the creation request sent by the cloud controller, saving it to the cloud database and the cloud storage, and running the application. When a node controller in one of the cars fails and causes an application failure on its node, the application template of its application will monitor the abnormality of the application, and then notify the cloud controller to schedule to other normal node controllers to recreate the application, thereby solving the node controller failure point problem in the vehicle-mounted cloud platform.

[0054] The data unit includes a cloud database 309, a cloud database backup module 310, and a cloud database monitoring module 311. The cloud database is deployed on the master and standby cloud controller nodes in a master-slave manner, and is distributed one-to-one according to the cloud controllers. It is mainly used to store vehicle-mounted cloud platform resource data, mainly including application templates, application instances, application images, storage volumes, network IP and other information; the backup module is mainly used for database backup after the cloud platform is completed in the application adaptation phase; the monitoring module is used to monitor the working status of the master and slave databases in real time and synchronize the master and slave database data. When the database status has abnormalities such as brain split, the monitoring module will automatically restore the database through the database information backed up by the backup module to keep the master and standby database information consistent, thereby ensuring the high availability of the cloud database.

[0055] The cloud database backup module is used to back up the database resources of the vehicle-mounted cloud platform after the application adaptation phase is completed, and is saved as backup_init.tar.gz; the cloud database monitoring module is used to synchronize the primary and standby database data in real time and monitor the database status. Specifically, the cloud storage monitoring module is used to monitor the status of the cloud storage in real time, and when an abnormality occurs in the primary and standby cloud storage, it automatically synchronizes the primary and standby cloud storage data to ensure that the cloud storage is available. The cloud database monitoring module is used to monitor the status of the cloud database in real time, and when a brain split abnormality occurs in the primary and standby cloud controllers, it restores the data through the backup file backup_init.tar.gz of the cloud database backup module to ensure the availability of the cloud database.

[0056] The storage unit includes cloud storage 312 and cloud storage monitoring module 313. The cloud storage is deployed on the master and standby cloud controllers in a master-slave manner and distributed one-to-one according to the cloud controllers to provide storage volume mounting for the application unit; the cloud storage monitoring module is used to synchronize the master and standby cloud storage and monitor the status of the cloud storage. When the cloud storage has abnormalities such as brain split, the monitoring module selects the master storage through the virtual IP management component, and then simultaneously transfers the data of the master storage to the slave end to keep the master and slave storage data consistent; on the premise that the storage data lost by the application can be tolerated, the high availability of the cloud storage is guaranteed.

[0057] Furthermore, when critical components of the cloud platform are abnormal under harsh environments such as weak electricity and weak network, the component status can be automatically determined and a self-recovery mechanism can be provided. Critical component abnormalities include: cloud database split-brain, cloud storage split-brain, cloud controller split-brain, node controller power failure, etc. The self-recovery mechanism includes: backing up the system configuration and application data when the cloud platform is online, rebuilding the environment when the component is abnormal through the backup data, self-recovery of component services after network partition, and self-recovery of application services. Rebuilding the abnormal component environment includes self-recovery of the primary and standby databases from backup data when the cloud database is split-brain, and self-recovery through data synchronization when the cloud storage is split-brain; self-recovery of component services after network partition includes independent operation of components in the primary and standby mode to provide services when the network is abnormal, and automatic recovery of the primary and standby mode to continue providing services after the network is restored; self-recovery of application services includes automatic creation when the application is not running, re-creation when the application is abnormal, and deletion of redundant applications when the same application is running on multiple nodes at the same time.

[0058] Figure 3 The diagram shows a method for deploying a high-availability system for a vehicle-mounted cloud platform application according to the present invention. Figure 3 As shown, the specific steps of this deployment method are as follows:

[0059] Step 1: After the cloud platform is deployed on the embedded board, the system enters maintenance mode. The vehicle-mounted cloud platform is used for vehicle-mounted application adaptation, image packaging, storage data creation, testing and debugging, etc. The cloud controller creates an application instance, configures and adapts the application, and ensures that the modules in the platform work properly.

[0060] Step 2: Based on the application requirements and resource configuration (such as images, storage volumes, network configuration, etc.), the cloud controller notifies the application self-maintenance component to create an application template, which contains the necessary configuration required by the application, such as resource requirements, deployment strategies, etc.

[0061] Step 3: The application self-maintenance component processes the creation of the application template, and the application template will be saved in the cloud database to ensure that the template can be called and used in the subsequent production mode.

[0062] Step 4: The vehicle-mounted cloud platform backs up all key data in the cloud database. The backup file will contain important resource data such as application templates, image information, storage volume configuration, etc. The backup file is named backup_init.tar.gz and stored in the primary and backup cloud database nodes. These backup data can be used for subsequent disaster recovery, environment rollback and other operations. After the cloud database backup is completed, switch the vehicle-mounted cloud platform from maintenance mode to production mode through the management console. During the switching process, the platform will ensure that all resources are ready and check the health status of all services and components. After switching to production mode, the cloud platform will start all components and functions required in production mode to ensure that all applications in production mode can run stably under strict production environment conditions.

[0063] Step 5: The system starts in production mode, and the application self-maintenance component notifies the cloud controller to automatically create a new application instance based on the application template saved in step 2.

[0064] Step 6: The cloud controller will schedule the corresponding node controller to create an application instance based on the specified resource requirements (such as image, CPU, memory, storage volume, etc.). The cloud platform will automatically allocate resources according to application requirements and ensure that the resource allocation of each application instance meets the standards of the production environment. The vehicle-mounted cloud platform will go online.

[0065] Step 7: During the online operation phase, the vehicle cloud platform continuously monitors the status of the corresponding modules through the application template, the active and standby cloud controller, the cloud storage monitoring module, and the cloud database monitoring module, thereby ensuring the high availability of the node application when the vehicle cloud platform is online. Figure 4 As shown:

[0066] a. When an application crashes due to storage, network, or node controller anomalies, the application template notifies the cloud controller to reschedule to a new node controller to create the application, ensuring high availability of the in-vehicle application under the premise that the cloud platform components are normal.

[0067] b. When the cloud controller malfunctions due to abnormalities such as network partitioning in the car, after the network is restored, the cloud platform restores the master-slave operation status of the cloud controller according to the pre-configured node priority. The node with the highest priority becomes the master cloud controller, thereby ensuring the high availability of the cloud controller.

[0068] c. When the car frequently loses points, loses network, or other abnormalities, causing cloud storage abnormalities, after the abnormality is recovered, the cloud storage monitoring module selects the primary storage through the virtual IP management component, and then synchronizes the data of the primary storage to the slave end, restoring the cloud storage status, thereby ensuring the availability of the cloud storage.

[0069] d. When abnormalities such as point loss and network loss occur, the cloud database may become abnormal, resulting in data inconsistency or database service unavailability. The database monitoring module will continuously monitor the status of the database, detect abnormal phenomena and record faults in a timely manner, and automatically restore the database status from the backup module (such as backup_init.tar.gz). During the recovery process, the system will perform a data consistency check to ensure that the data restored from the backup is complete and correct.

[0070] The above-mentioned high-availability system and deployment method of the vehicle-mounted cloud platform application have the following technical effects:

[0071] 1) Through the high availability design of the above-mentioned application unit, control unit, execution unit, storage unit and data unit, the cloud platform realizes automatic maintenance of vehicle applications during the application launch phase, ensuring the high availability and high reliability of vehicle applications.

[0072] The debugging phase and the online operation phase of the cloud platform vehicle application are separated by design, so as to ensure the purity of the cloud platform environment and the certainty of resources during the online operation phase.

[0073] 2) Through the distributed, high-concurrency, and lightweight features of Erlang at the language level, lightweight and simple deployment of cloud controllers and node controllers can be achieved, while ensuring high availability of cloud controllers and node controllers.

[0074] 3) During the online operation phase of the cloud platform, application templates managed by application template components are used to achieve one-to-one creation and monitoring of applications, which can ensure high availability of vehicle-mounted applications under the premise of a normal cloud platform environment.

[0075] 4) When abnormalities such as power outages and network disconnections cause brain splits in key components of the cloud platform, the self-recovery capabilities of each component ensure the high reliability of the cloud platform, thereby ensuring the high availability of in-vehicle applications.

[0076] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Therefore, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A high-availability system for vehicle-mounted cloud platform applications, characterized in that: The system includes: an application unit, a control unit, an execution unit, a storage unit and a data unit. The control unit includes a cloud controller, which is distributed in two adjacent carriages in a master-slave deployment mode, and the cloud controller is used to send resource requests, receive application creation requests and schedule node controllers to process the creation of the application; The application unit includes an application self-maintenance component, and the application self-maintenance component is used to accept a creation request of an application template issued by the cloud controller; The storage unit includes cloud storage and a cloud storage monitoring module. The cloud storage is distributed one-to-one according to the cloud controller in a primary-backup manner and is used to provide storage volume mounting for the application unit; the cloud storage monitoring module is used to synchronize the primary and backup cloud storage and monitor the status of the cloud storage; The data unit includes a cloud database, a cloud database backup module, and a cloud database monitoring module. The primary and backup cloud databases are distributed one-to-one according to the cloud controller in a primary-backup manner, and are mainly used to store resource data of the vehicle-mounted cloud platform; the cloud database backup module is used to back up the database resources of the vehicle-mounted cloud platform after the application adaptation phase is completed, and is saved as backup_init.tar.gz; the cloud database monitoring module is used to synchronize the primary and backup database data in real time and monitor the database status; The execution unit includes a plurality of node controllers, which are distributed in each carriage. The node controllers are mainly responsible for processing the creation request sent by the cloud controller and saving it to the cloud database and the cloud storage.

2. The vehicle-mounted cloud platform application high availability system according to claim 1, characterized in that: The application template saves the application information required by the vehicle cloud to the database during the application adaptation phase, including the image, storage volume, and IP resources required by the application; During the application launch phase, the application template automatically creates and monitors the application. After an application failure, the application template actively sends an application creation request to the cloud controller.

3. The vehicle-mounted cloud platform application high availability system according to claim 1, characterized in that: The cloud controller realizes the master-slave capability through the distributed mechanism of Erlang; when the master controller is running, the standby controller stops; when the controlled device cannot connect to the master controller through the network heartbeat, the standby controller runs, thereby ensuring the master-slave high availability of the cloud controller.

4. The vehicle-mounted cloud platform application high availability system according to claim 1, characterized in that: The cloud storage monitoring module is used to monitor the status of the cloud storage in real time, and automatically synchronize the primary and backup cloud storage data when an abnormality occurs in the primary and backup cloud storage to ensure that the cloud storage is available.

5. The vehicle-mounted cloud platform application high availability system according to claim 1, characterized in that: The cloud database monitoring module is used to monitor the status of the cloud database in real time, and when a brain split occurs in the active and standby cloud controllers, restore data through the backup file backup_init.tar.gz of the cloud database backup module to ensure the availability of the cloud database.

6. A deployment method for a high-availability system for an in-vehicle cloud platform application according to any one of claims 1 to 5, characterized in that: The specific steps of the deployment method are as follows: After completing the cloud platform deployment on the embedded board and completing the linkage work between the vehicle-mounted application and the vehicle-mounted device in the adaptation phase, physical resources are created on the node controller and data is saved to the cloud storage and cloud database; Sending a request to create the application template to the application self-maintenance component through the cloud controller; The application self-maintenance component binds and creates the application template according to the logical resource information and saves it in the cloud database; After the application template is successfully created, the cloud database backup operation is performed to back up the file backup_init.tar.gz, and the vehicle-mounted cloud platform is ready to enter the online operation stage; The application template notifies the cloud controller to create an application. The cloud controller processes the creation of the application by scheduling it to the node controller, and saves the application resource information in the cloud database, and enters the online operation stage.

7. The method for deploying a high-availability system for a vehicle-mounted cloud platform application according to claim 6, characterized in that: When the application crashes due to an exception, the application template notifies the cloud controller to reschedule to a new node controller to create the application; When an abnormality occurs in the cloud controller, after the network is restored, the vehicle-mounted cloud platform restores the active and standby operation status of the cloud controller according to the node priority configured in advance, and the node with the highest priority becomes the active cloud controller; When an exception occurs in the cloud storage, after the exception is recovered, the cloud storage monitoring module selects the primary storage through the virtual IP management component, and then synchronizes the data of the primary storage to the backup storage to restore the cloud storage state; When the cloud data is abnormal, after the abnormality is recovered, the cloud database monitoring module restores the cloud database state from the backup file backup_init.tar.gz data.

8. The method for deploying a high-availability system for a vehicle-mounted cloud platform application as claimed in claim 7, characterized in that: The physical resources created by the node controller include images, storage volumes, and network data; the application self-maintenance component binds and creates the application template according to logical resource information, and the logical resource information includes images, storage volumes, and network IP addresses.

9. The method for deploying a high-availability system for a vehicle-mounted cloud platform application as claimed in claim 8, characterized in that: The application anomalies include downtime caused by storage module, network or node controller anomalies; the cloud controller anomalies include network partition anomalies causing a split-brain anomaly in the cloud controller; the cloud storage anomalies include frequent network dropouts and network disconnection anomalies causing a split-brain anomaly in the cloud storage.

Images