A disk storage capacity prediction method based on a time sequence model

Abstract

The invention discloses a disk storage capacity prediction method based on a time sequence model, which comprises the following steps of (1) establishing a database for storing disk capacity use conditions, monitoring the operation of a storage system, and providing data required for prediction; (2) reading a database, obtaining historical data, carrying out data processing, constructing a prediction model based on a time sequence model, and carrying out prediction; and (3) comparing the capacity prediction result with the residual capacity, evaluating and judging whether the capacity prediction result accords with a preset threshold value, judging whether to send out early warning, and reminding operation and maintenance personnel to maintain the disk system. According to the intelligentprediction technology for the disk capacity, different time sequence models are selected according to historical data characteristics, the limitation of a prediction result of a single method is avoided, the storage prediction requirements of different disk systems are met, and the very strong adaptability and universality are achieved. According to the method, the disk capacity can be predicted more accurately in an actual test, and the labor and the financial cost of system operation and maintenance are reduced.

Description

technical field [0001] The invention relates to disk monitoring and disk capacity prediction technology, in particular to a method for predicting disk storage capacity based on a time series model. Background technique [0002] With the rapid development of Internet big data, large-scale data centers have increasingly become the development needs of State Grid Corporation. This poses a huge challenge to disk management in data centers. Disk management in a data center with PB-level storage is already very difficult, not to mention that the current development trend is that larger data centers are gradually emerging. For example, the monitoring of a data center with 1000PB storage cannot be satisfied by the traditional disk monitoring method. Even if the traditional monitoring method can monitor the status of the disk, it cannot guarantee the full and efficient use of the disk capacity, often causing some disks to Idle for a long time, while other disk resources are exhauste...

AI Technical Summary

Problems solved by technology

[0005] When using the disk capacity prediction method based on historical monitoring data, because the disk capacity not only depends on the historical state, but more importantly, the current system operating state, and because the historical data has not been cleaned and reconstructed, the prediction results may be relatively poor. big deviation

Images