Large-scale automatic cluster performance testing method and system based on cloud service

Abstract

The invention discloses a large-scale automatic cluster performance testing method and system based on cloud service. The method comprises the following steps: generating computing nodes in batches on a physical server based on an operating system provided with testing software; specifying the test software of one computing node in the generated computing nodes as a server, and taking the test software of the other computing nodes as clients; and issuing a test program and a test task based on the test program to a specified client through the server, and synchronously acquiring resource occupancy states of the client and a physical server / physical client during execution of the test task by each client. According to the method, respective performance boundaries which can be reached by a cloud computing solution under the conditions of the same hardware and different hardware can be defined, respective performance boundaries which can be reached by different cloud computing solutions under the same hardware can be defined, testing efficiency is high, and test errors are small.

Description

technical field [0001] The invention relates to a cloud service-based large-scale cluster performance testing technology, in particular to a cloud service-based large-scale cluster performance automatic testing method and system. Background technique [0002] Large-scale clusters based on cloud services are currently the basic architecture for providing cloud services. For example, virtual machines are provided on physical servers as computing nodes, and physical clients log in to computing nodes to realize various remote operations including cloud desktops or calculate. At present, large-scale clusters based on cloud services have the following problems: it is difficult to define the upper limit resources that can be run by virtual machines in large-scale clusters based on cloud services, and it is impossible to define the computing resources that need to be allocated for certain operations inside virtual machines. The size of the network resources required for the remote ...

AI Technical Summary

Problems solved by technology

At present, large-scale clusters based on cloud services have the following problems: it is difficult to define the upper limit resources that can be run by virtual machines in large-scale clusters based on cloud services, and it is impossible to define the computing resources that need to be allocated for certain operations inside virtual machines. The size of the network resources required for the remote operation of the virtual machine cannot be accurately defined, and the maximum performance that the hardware of the cloud server can achieve, as well as the impact of the hardware usage time, can not effectively perceive the performance degradation presented by aging.

For the testing of large-scale clusters based on cloud services, most of the comparison tests are carried out by testers. Each tester has his own test methods and test habits. The reflection time and test monitoring time will cause human interference errors to the test results. Moreover, different test scripts currently on the market cannot be written for multiple platforms at one time for execution. For example, test scripts written on win7 cannot be tested on win10, win2008 or Linux

Images