Qianxun SI Serves 1.1 Billion Users At The Same Time, Relying On It!
Qianxun SI's high-precision spatial services have served more than 1.1 billion people worldwide, achieving "unlimited" service access capability through "limited" ground-based augmentation stations, relying on high concurrency.
Distributed Computing To Handle "Heavy Traffic" Comfortably
What is high concurrency? Just like in E-commerce promotion or Chinese New Year ticket rush, when encountering a large number of users with fast access, if it does not have good high concurrency processing capability, it will trigger service interruption and cause bad experience to users. And for devices such as drones and self-driving cars in operation, service interruptions can even create security risks.
"The data calculation and related links involved in each user visit are much more complicated than the order system of an e-commerce platform. From the satellite in the sky to the reference station, to data processing, data broadcasting, and finally providing high-precision positioning services to global users, any problem in the whole chain will lead to unavailability of the service." Qianxun SI's senior JAVA development expert introduced.
To address such a situation, Qianxun SI is designed with a distributed service architecture to achieve horizontal scalability. To understand it simply, it is like a cafe that suddenly receives a large wave of customers, and to avoid customers queuing, it can reduce the pressure of single-line queuing by adding service windows.
Qianxun SI's distributed architecture model is divided into four levels, which solve four problems of distributed computing, rapid expansion, centralized processing, and load balancing, allowing the system to have the ability to "Turn a big complex problem into many small simple problems". At present, the number of computer clusters of its distribution platform exceeds one thousand, and the data processing volume reaches PB level, far exceeding the industry level.
Among them, SpaceX is a high-performance distributed application framework independently developed by Qianxun SI, which is the basic component of business and product back-end services, solving the problems of limited access to traditional mapping industry sites, insufficient service stability, and insufficient real-time broadcasting.
SpaceX is like a "manager ", coordinating the tasks. The system integrates the SpaceX technology framework to divide a large computing task into several smaller tasks and assign them to each computer. The cluster of computers improves efficiency through parallel processing and accomplishes large-scale information processing tasks under unified management and control. The national BeiDou ground-based augmentation system "One Network" is solved by SpaceX framework, divided into thousands of small network tasks, distributed in hundreds of machines to solve at the same time, and finally produce hundreds of thousands of grid data nationwide.
SpaceX technology framework quickly realizes a cluster architecture of distributed processing system with high throughput, low latency and high availability in the application system through task scheduling and task management, computer health status checking and switch in seconds, load balancing, routing and grayscale upgrade.
Currently, SpaceX is still evolving to achieve the adaptation of more data protocol access, more decomposition unit data broadcast, to ensure the efficient and stable operation of the broadcast link.
Operations & Drills To Ensure Uninterrupted Service
If the distributed architecture solves the problem of how to support hundreds of millions of user services, then the next problem to be solved is how to ensure the high availability of services, i.e., 7*24 hours uninterrupted service.
For several years, Qianxun SI has continuously and stably served tens of thousands of drones, 100,000 measurement and various engineering devices, millions of smart cars, and hundreds of millions of cell phones and IoT devices. This is backed by 7*24 hours non-stop online operation and maintenance monitoring, as well as effective failure drills. Its distribution platform has been rehearsed for disaster-tolerant failures, and the high availability of the entire link is guaranteed through repeated automated testing by continuously accumulating experience in testing under various extreme scenarios.
Even in the event of a major geographical disaster, Qianxun SI's distribution platform can guarantee uninterrupted service operation through its network redundancy design and switching mechanism, which allows the system to automatically complete switching when equipment or lines have problems.
When the industry ushers in a new growth phase, technology is the important driving force behind it. At present, the high concurrency capacity of Qianxun SI has reached one million per second, a figure that far exceeds that of the same industry and is among the best even in the Internet industry. Qianxun SI will continue to carry out independent and controllable technology research to provide a solid technical base for high-precision services to be used by the public on a larger scale.