Method and device for testing and displaying storage performance prediction function

Abstract

The invention discloses a method and device for testing and displaying a storage performance prediction function. The method comprises the steps of obtaining performance historical data based on a storage cluster; acquiring data of a first preset number of days from the performance historical data as training data; acquiring data of a second predetermined number of days after the first predetermined number of days from the performance historical data as test data; performing prediction based on the training data through a performance prediction engine, and obtaining prediction data of a second predetermined number of days after the first predetermined number of days; and generating a performance historical data curve, a test data curve and a prediction data curve, and comparing the test data with the prediction data to obtain a prediction error of the performance index. The invention can be used for displaying the accuracy of the prediction result, a user can conveniently and visually experience the effect of the performance prediction function, and the problems that the storage performance prediction function cannot be evaluated and tested and manual summarization and calculation of a large amount of data are tedious and time-consuming are solved.

Description

technical field [0001] The invention relates to the field of storage technology, in particular to a method and device for testing and displaying storage performance prediction functions. Background technique [0002] Performance is one of the most important capabilities provided by a storage system and is critical to the smooth operation and maintenance of a data center. As the base of the IO stack of the IT system, if the storage has performance problems, it will cause a series of problems in the upper application. Generally speaking, these problems are very difficult to locate (full-stack performance analysis is required). Therefore, for the operation and maintenance personnel of the data center, a major challenge is how to ensure that the storage performance meets the needs of the business system and avoid performance problems. The premise of solving this problem is to have a clear understanding of the performance requirements of the storage system and to be able to accu...

AI Technical Summary

Problems solved by technology

This method is a waste of time (at least 2 months), and laborious (data aggregation, collation, calculation), and the final result is only an error value, which cannot intuitively allow users to see the gap between the two

In the same way, when testing the storage performance prediction function, you will also face the same problem

Images