System and method for quantizing and calculating service level of throughput of cloud system

Abstract

The invention discloses a system for quantizing and calculating a service level of a throughput of a cloud system, which comprises a data acquiring module, a data screening module, a data editing module, a data calculating module and a data evaluating module. The data acquiring module is used for acquiring TPS data of the cloud system; the data screening module is used for screening and removing head end data and tail end data in the TPS data and transferring residual active TPS data to the data editing module; the data editing module is used for outputting a TPS average value and a TPS standard deviation to the data calculating module; the data calculating module calculates a TPS fluctuation value of the TPS data and outputs the TPS fluctuation value to the data evaluating module; and thedata evaluating module carries out rating on the throughput of the cloud system and outputs an evaluation result. The system disclosed by the invention cannot excessively depend on a performance testing tool, data acquisition and testing can be automatically carried out, and finally, evaluations on fluctuation of throughputs of different interfaces and functions are uniformly shown with percentage values so as to facilitate a tester directly acquire the data and know the throughput of the cloud system.

Description

technical field [0001] The invention relates to the technical field of performance testing, in particular to a system and method for quantifying and calculating the service level of the throughput capability of a cloud system. Background technique [0002] In the field of performance testing, the throughput of interfaces and functions is the most critical indicator for evaluating the performance of cloud systems. On the one hand, insufficient throughput may cause the cloud system to take too long to respond and even crash, seriously affecting user experience. On the other hand, throughput Excessive fluctuations may cause the cloud system to respond unstable, and end users may access the system at different speeds, which also seriously affects the user experience. Therefore, the evaluation of the volatility of the throughput is as important as the measurement of the average value of the throughput. [0003] Currently, there are two methods for displaying the volatility of th...

AI Technical Summary

Problems solved by technology

These two methods have their limitations. Although the former can intuitively display volatility, it cannot make digital evaluations. The latter can make digital evaluations, but the value varies with interfaces and functions. Large values It does not necessarily mean that the volatility is large, and the small data does not mean that the code volatility is small, so it is impossible to make a unified evaluation based on the value

Images