System and method for improving server usage based on anomaly detection
A management server includes linear regression circuitry to receive the data as input data, based on the input data, the linear regression circuitry generates a linear regression line for the input data. Gradient circuitry analyzes angles between two points of data within the input data. In response to a first angle between a first point of data and a second point of data being greater than a threshold angle, the gradient circuitry determines that an interval of data including the first and second points of data is an abrupt interval, removes first data within the abrupt interval from the input data, and updates the linear regression line accordingly. Support vector machine circuitry receives the input data, classifies outliers in the input data, and removes the outliers. The linear regression line is updated based on the removal of the outliers.
1. An information handling system, comprising:
a first server to collect data associated with a metric of the first server; and
a management server to receive the data from the first server, the management server including:
linear regression circuitry to receive the data as input data, and based on the input data the linear regression circuitry to generate a linear regression line for the input data, wherein the linear regression line indicates a pattern within the input data;
gradient circuitry to analyze angles between two points of data within the input data; in response to a first angle between a first point of data and a second point of data being greater than a threshold angle, to determine that an interval of data including the first and second points of data is an abrupt interval; and to remove first data within the abrupt interval from the input data to update the linear regression line; and
support vector machine circuitry to receive the input data, to classify outliers in the input data, and to remove the outliers, wherein the linear regression line is updated based on the removal of the outliers wherein during the monitoring phase and in response to the detection of the anomaly: the management server to enter into an anomaly detection period; in response to an amount of time within the anomaly detection period being greater than a predetermined amount of time, entering, by the processor, into a recovery management period; and during the recovery management period, performing an operation to increase storage capacity within the information handling system.
2. The information handling system of claim 1 , wherein during the monitoring phase: the linear regression circuitry to determine whether the input data is within a predetermined percentage of the linear regression line, in response to the input data not being within the predetermined percentage of the linear regression line, the gradient circuitry to determine whether a second angle between a second point of data and a third point of data is greater than the threshold angle, in response to the second angle being greater than the threshold angle, the support vector machine circuitry to determine an intensity of change within the input data received after the determination that the second angle is greater than the threshold angle, and in response to the intensity of change being greater than a threshold, the support vector machine circuitry to detect the anomaly within the input data.
3. The information handling system of claim 2 , wherein during the monitoring phase and in response to the detection of the anomaly: the management server to enter remain in the anomaly detection period while the intensity of change remains greater than the threshold, to determine whether an amount of time within the anomaly detection period is greater than a predetermined amount of time.
4. The information handling system of claim 3 , further comprising:
a display device to communicate with the management server, in response to the amount of time within the anomaly detection period being greater than the predetermined amount of time, the display device to provide a warning message to a user of the information handling system.
5. The information handling system of claim 2 , wherein the detection of the anomaly indicates that a rate of storage capacity usage of a disk space within the information handling system has increased to a level that will exceed a storage capacity of the disk space.
6. The information handling system of claim 1 , wherein the angle between the two points of data is a degree of change between the two points.
7. A method, comprising:
receiving, at a processor of an information handling system, input data for a metric of the information handling system, wherein the input data is periodically received during both a training phase and a monitoring phase; and
during the training phase:
training, by the processor, linear regression circuitry based on the input data;
generating, based on the training of the linear regression circuitry, a linear regression line for the input data, wherein the linear regression line indicates a pattern within the input data;
executing, by the processor, gradient circuitry to analyze angles between two points of data within the input data;
in response to a first angle between a first point of data and a second point of data being greater than a threshold angle, determining, by the gradient circuitry, that an interval of data including the first and second points of data is an abrupt interval;
removing first data within the abrupt interval from the input data;
executing, by the processor, support vector machine circuitry to classify outliers in the input data;
removing the outliers from the input data; and
updating, the removing of the first data and the removing of the outliers, the linear regression line for the input data; and
during the monitoring phase:
in response to a detection of an anomaly, entering, by the processor, into an anomaly detection period;
in response to an amount of time within the anomaly detection period being greater than a predetermined amount of time, entering, by the processor, into a recovery management period; and
during the recovery management period, performing an operation to increase storage capacity within the information handling system.
8. The method of claim 7 , further comprising:
during the monitoring phase:
determining, by the linear regression circuitry, whether the input data is within a predetermined percentage of the linear regression line;
in response to the input data not being within the predetermined percentage of the linear regression line, determining, by the gradient circuitry, whether a second angle between a second point of data and a third point of data is greater than the threshold angle;
in response to the second angle being greater than the threshold angle, determining, by the support vector machine circuitry, an intensity of change within the input data received after the determination that the second angle is greater than the threshold angle; and
in response to the intensity of change being greater than a threshold, detecting the anomaly within the input data.
9. The method of claim 8 , further comprising:
remaining in the anomaly detection period while the intensity of change remains greater than the threshold; and
determining, by the processor, whether the amount of time within the anomaly detection period is greater than the predetermined amount of time.
10. The method of claim 9 , further comprising:
in response to the amount of time within the anomaly detection period being greater than the predetermined amount of time, providing, by a display device of the information handling system, a warning message to a user of the information handling system.
11. The method of claim 8 , wherein the detecting of the anomaly indicates that a rate of storage capacity usage of a disk space within the information handling system has increased to a level that will exceed a storage capacity of the disk space.
12. The method of claim 7 , wherein the angle between the two points of data is a degree of change between the two points.
13. A method, comprising:
receiving, at a processor of an information handling system, input data for a metric of the information handling system, wherein the input data is periodically received during both a training phase and a monitoring phase;
during the training phase:
training, by the processor, linear regression circuitry based on the input data;
generating, based on the training of the linear regression circuitry, a linear regression line for the input data, wherein the linear regression line indicates a pattern within the input data;
executing, by the processor, gradient circuitry to analyze angles between two points of data within the input data;
in response to a first angle between a first point of data and a second point of data being greater than a threshold angle, determining, by the gradient circuitry, that an interval of data including the first and second points of data is an abrupt interval;
removing first data within the abrupt interval from the input data;
executing, by the processor, support vector machine circuitry to classify outliers in the input data;
removing the outliers from the input data; and
updating, based on the removing of the first data and the removing of the outliers, the linear regression line for the input data; and
during the monitoring phase:
determining, by the linear regression circuitry, the gradient circuitry, and the support vector machine, whether an anomaly is detected within the input data;
in response to the anomaly being detected, entering into an anomaly detection period;
in response to the anomaly detection period, providing, by a display device of the information handling system, a warning message to a user of the information handling system;
in response to a detection of an anomaly, entering, by the processor, into an anomaly detection period;
in response to an amount of time within the anomaly detection period being greater than a predetermined amount of time, entering, by the processor, into a recovery management period; and
during the recovery management period, performing an operation to increase storage capacity within the information handling system.
14. The method of claim 13 , wherein the determining of whether the anomaly is detected within the input data further comprises:
determining, by the linear regression circuitry, whether the input data is within a predetermined percentage of the linear regression line;
in response to the input data not being within the predetermined percentage of the linear regression line, determining, by the gradient circuitry, whether a second angle between a second point of data and a third point of data is greater than the threshold angle;
in response to the second angle being greater than the threshold angle, determining, by the support vector machine circuitry, an intensity of change within the input data received after the determination that the second angle is greater than the threshold angle; and
in response to the intensity of change being greater than a threshold, detecting the anomaly within the input data.
15. The method of claim 13 , further comprising:
remaining in the anomaly detection period while the intensity of change remains greater than the threshold;
determining, by the processor, whether an amount of time within the anomaly detection period is greater than a predetermined amount of time;
in response to the amount of time within the anomaly detection period being greater than the predetermined amount of time, entering, by the processor, into a recovery management period.
16. The method of claim 13 , wherein the detecting of the anomaly indicates that a rate of storage capacity usage of a disk space within the information handling system has increased to a level that will exceed a storage capacity of the disk space.
17. The method of claim 13 , wherein the angle between the two points of data is a degree of change between the two points.