Fault prediction method and system for storage device and related device
A technology for fault prediction and storage equipment, applied in error detection/correction, instrumentation, computing, etc., can solve problems such as difficult to apply production environment, low model accuracy and recall rate, performance degradation, etc., to improve the accuracy of fault prediction. Effect
Active Publication Date: 2020-01-03
INSPUR SUZHOU INTELLIGENT TECH CO LTD
View PDF2 Cites 5 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0002]Aiming at the problem of performance degradation or even data loss caused by storage device failures in large-scale data center storage systems, many researches on fault prediction methods have emerged. Because there are fewer bad storage devices and more good storage devices, The ratio of positive and negative samples is unbalanced (about 3:100), and the number of positive samples available is small. In addition, experience shows that simply relying on the current SMART characteristic value of the storage device cannot effectively judge whether the storage device is faulty or not, and the change of the SMART value It can better reflect the potential failure of the storage device, so the accuracy and recall of the model constructed by the above method are low, and it is difficult to apply in a normal production environment
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0037] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.
[0038] Please refer to figure 1 , figure 1 It is a flow chart of a storage device failure prediction method provided in an embodiment of the present application, and the method includes:
[0039] S101: Obtain the latest status data of the storage device;
[0040] It should be noted that, by default, the binary classification model needs to be gener...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides a fault prediction method for a storage device. The method comprises the steps of obtaining latest state data of the storage equipment; performing time sequence feature extraction of a preset feature extraction window on the latest state data based on a tsfresh library to obtain a time sequence feature value; obtaining a test characteristic value of the storage device according to the time sequence characteristic value and an original SMART RAW value; and inputting the test characteristic value into a binary classification model to obtain a fault prediction result. According to statistical analysis of a historical bad storage device, a data set is labeled again based on a threshold value, so that the number of positive samples is increased. Based on tsfresh time sequence feature extraction, dynamic and static features of SMART features are combined, the limitation of dichotomy is broken through, and the fault prediction accuracy of the storage device can be improved. The invention further provides a fault prediction system for the storage device, a computer readable storage medium and a fault prediction terminal, which have the above beneficial effects.
Description
technical field [0001] The present application relates to the field of storage devices, and in particular to a storage device failure prediction method, system and related devices. Background technique [0002] Aiming at the problem of performance degradation or even data loss caused by storage device failures in large-scale data center storage systems, many studies on fault prediction methods have emerged. Due to the fact that there are few bad storage devices and many good storage devices, and the ratio of positive and negative samples is unbalanced (about 3:100), the available The number of positive samples is small. In addition, experience shows that relying solely on the current SMART characteristic value of the storage device cannot effectively judge whether the storage device is faulty or not. The change of the SMART value can better reflect the potential fault of the storage device. Therefore, the above method is used to construct The accuracy and recall of the model...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F11/34
CPCG06F11/3409
Inventor 王团结苏楠李辉
Owner INSPUR SUZHOU INTELLIGENT TECH CO LTD