The ** company focuses on the commercial application of autonomous driving. Based on over ten years of technological accumulation and more than three
years of commercial operation experience, it leverages the advantages of the industrial chain by utilizing industry-leading full-stack self-developed
intelligent driving technology. The company concentrates on specific scenarios such as ports, mining areas, parks, and trunk lines. It provides
intelligent driving solutions and intelligent transport capacity services across various fields and scenarios. The development process has
successfully transitioned to microservices and containerization, currently operating within a high-availability environment based on a single physical
server.
This setup is essential for ensuring business continuity and stability in the all-day autonomous operation of large vehicles at the
port, necessitating robust service capabilities.
The IT operation and maintenance team has limited energy and aims to achieve simple delivery,
intuitive operation, and comprehensive monitoring capabilities, including alarm management, log viewing, and error debugging.
The project’s
online cycle is short, iteration is rapid, and swift delivery is required. A hyper-converged cluster consists of three servers and two 10-gigabit
switches. MySQL and Redis services run directly on virtual machines as stateful services.
A hyper-converged cluster employs six VM groups to
form an K8s cluster. All microservice components operate as containers within the K8s cluster and exclusively access MySQL and Redis services in the
same cluster.
To meet the requirements of the active-active switchover, services in the cluster utilize Nginx as the access portal to facilitate
health checks and traffic switching when faults occur.
Architecture Design： Design a hyper-converged architecture, encompassing servers, storage,network configuration, and K8s clustering.
Database Design： Determine the deployment scheme of MySQL and Redis to ensure high availability and consistency of data.
Hardware Deployment： Install and configure servers and storage devices at each site, including primary and secondary equipment rooms.
Virtualization Configuration： Configure a hypervisor at each site and deploy service VMs and K8s clusters.
Database Configuration： Install and configure MySQL and Redis databases to ensure their high availability.
Data Synchronization： Configure the primary/secondary synchronization and asynchronous replication mechanisms of MySQL to ensure real-time data synchronization between different sites.
Health Check Configuration： Configure the health check functionality of Nginx to ensure the high availability of the system.
Failover Configuration： Establish a failover mechanism to ensure that the system can automatically switch to backup resources in the event of a fault.
Rapid Deployment, High availability, superior performance, unified management, and simplified operation and maintenance.
The K8S cluster enables the transformation of microservice applications in Docker into a K8S cluster.
Enhanced stability： Multi-copy service deployment facilitates second-level traffic switching.It also supports automated health monitoring and recovery to ensure uninterrupted business operations.
Accelerated Platform Iteration： Standardization of the development and launch processes has significantly expedited the FMS platform's iteration,achieving minute-level deployment from the original hourly launch.
Simplified O&M： Storage,virtualization,and K8S management are integrated into a single interface,offering diversified monitoring capabilities.
Cost saving： A minimum three-node cluster can be established,ensuring high availability and comprehensive platform functionalities.