Real-time cloud platform management monitoring system

Abstract

The invention provides a real-time cloud platform management monitoring system. The real-time cloud platform management monitoring system comprises a virtual machine, a resource virtualization layer, a real-time module and an interrupt card, wherein the real-time module improves the real-time performance of the system by putting real-time preempt patches on the resource virtualization layer and a virtual machine operating system layer; the interrupt card module receives an external signal, triggers hardware interrupt and requests a system kernel to process an event; and the interrupt card accesses a physical resource layer, transmits data to the virtual machine in an unvarnished manner and is used with strong real-time application. According to the method, a real-time patch, a resource virtualization layer and a virtual machine operating system layer are used for compiling and installing a real-time preempt patch (RT-Preempt Patch), shortens the interrupt response time, and improves the real-time performance of the system; the interrupt card is used, and strong real-time performance is achieved; and the physical resource layer uses interrupt card equipment, so that the real-time performance of the system is further improved, and the strong real-time application requirement is met.

Description

technical field [0001] The invention relates to the technical field of cloud platform management and monitoring, in particular to a real-time cloud platform management and monitoring system. Background technique [0002] At present, with the development of hardware performance and virtualization technology, various cloud management platforms based on virtualization technology (such as VMWare, Kvm, Xen, etc.) have emerged and are applied in various fields, such as artificial intelligence, image recognition, security, etc. Surveillance, smart transportation, etc. However, when a certain type of application is deployed on the cloud management platform, in addition to traditional resource allocation, there is also a high demand for real-time performance of resource scheduling. This type of application includes autonomous driving and in-memory databases. In order to meet the real-time performance requirements of such applications, some methods have emerged, such as "A Real-time ...

AI Technical Summary

Problems solved by technology

Comparison of technical points: 1. Low real-time performance: The scheme proposed in the above-mentioned document improves the average efficiency of super-call by using real-time super-call batch execution algorithm or cross-core super-call asynchronous execution algorithm, thereby achieving the purpose of improving the real-time performance of the system. The interrupt response time is not fundamentally reduced, and the real-time performance is low

2. Real-time performance cannot be guaranteed: the solution proposed in the above document only achieves real-time goals by improving efficiency, and cannot guarantee the lowest interrupt response time

Comparison of technical points: 1. Poor versatility: the solution proposed by the above patents uses a single-core microcontroller to realize real-time virtualization, which can meet the requirements of high real-time applications. However, for applications with general real-time performance requirements, this solution is adopted. The cost is too high and the versatility is poor

2. Single function: the solution proposed by the above patent only proposes an integrated virtualization system with a single function

[0005] For the prior art mentioned above, the inventor believes that the real-time performance of the above-mentioned method is low, and the real-time performance and general performance of the system cannot be guaranteed

Images