aSAN analyzes a large amount of real business scenario data, the underlying algorithm library integrates a variety of machine learning and deep learning algorithms, and independently developed the storage forecasting module, including bad sector prediction, capacity prediction and SSD life prediction functions.

1. Storage Capacity

aSAN can dynamically predict the capacity growth trend in the next 90 days based on the capacity usage of customer clusters. In the capacity forecast interface, users can switch to view the bare capacity of different storage volumes, the actual capacity used and the dynamic forecast curve, and the user is prompted in the section surface that the capacity used is expected to reach the capacity alarm threshold (90%) in xx days.

2. Bad Sectors

By collecting and analyzing the SMART data, performance parameters and hard disk log information of a large number of customers' real use scenarios, combined with advanced algorithm training models, Sangfor has independently developed a highly accurate bad sector forecasting function, and the aSAN bad sector forecasting accuracy rate is up to 95% or more through a large number of different business scenarios.

aSAN can dynamically sense which hard disks in the cluster are most likely to have bad sectors according to the upper business pressure and cluster hard disk usage, and prioritize them for bad sector scanning, which can shorten the actual window for hard disk bad sectors to be found from a week or even a month to a day, and combine aSAN bad sector scanning and repair to form a closed-loop processing of hard disk bad sectors, which greatly shortens the risk period of data being in "single copy" due to hard disk bad sectors.

In the Bad Sectors interface, users can see the hard disk bad sectors forecast results, actual scan results and bad sector repair, and at the same time will also be based on the number of hard disk bad sector risk assessment of the hard disk, for the history of a high number of bad sector hard disk alarm, prompting users to replace the hard disk as soon as possible to prevent data loss due to excessive number of bad channels resulting in hard disk damage.

3. SSD Lifetime

aSAN collects and analyzes the SSD hard disk IO data in the cluster, calculates the remaining life of SSD hard disk, and displays the expected remaining available time of SSD in combination with the upper business pressure, and divides the three life levels of healthy, risk and high risk according to the prediction result, prompting users to replace the SSD hard disk in the cluster with critical life in time.

High Risk: Remaining life <= 10%
Medium Risk: Remaining life <= 15%
Healthy: Remaining life > 15%