Systems and methods for detecting behavioral features of software application deployments for automated deployment management
First Claim
Patent Images
1. A system for monitoring and analyzing a deployment of a software application, comprising:
- a monitor for gathering time-variant state metric data from the deployment, the monitor associated with a computer system that runs components of the software application, the monitor configured to gather state metric data for at least one deployment state metric from the computer system; and
a non-transitory computer-readable medium that stores a numerical chart feature detector that, when executed by a processor, analyzes a numerical telemetry signal that comprises state metric data gathered by the monitor, said numerical chart feature detector configured to use a threshold and a prescribed sustain period to detect the presence of a feature based on the numerical telemetry signal, the feature representing a condition or behavior of the software application that is known to be associated with a potential problem of the deployment of the software application;
wherein the threshold comprises a zero-confidence level representing substantially zero confidence in the existence of a feature, a trigger-confidence level representing a prescribed intermediate confidence in the existence of a feature, and a full-confidence level representing a substantially maximum confidence in the existence of a feature;
wherein the numerical chart feature detector is configured to detect a feature when the telemetry signal exceeds the trigger-confidence level and remains in excess of the zero-confidence level subsequently throughout the prescribed sustain period immediately after exceeding the trigger-confidence level;
wherein an occurrence of the feature is represented by at least a name, confidence value, and time of occurrence.
9 Assignments
0 Petitions
Accused Products
Abstract
A system for monitoring and analyzing a deployment of a software application is disclosed. The system includes a monitoring component that collects streams of data values of time-variant state metrics associated with the deployment, and a feature detector that analyzes the streams to identify features known to be associated with potential problems. Detected features can include deviations of a signal from a normalcy band, a rising or descending trend in signal behavior, deviations from seasonal or cyclical normalcy bands, and the like.
-
Citations
46 Claims
-
1. A system for monitoring and analyzing a deployment of a software application, comprising:
-
a monitor for gathering time-variant state metric data from the deployment, the monitor associated with a computer system that runs components of the software application, the monitor configured to gather state metric data for at least one deployment state metric from the computer system; and a non-transitory computer-readable medium that stores a numerical chart feature detector that, when executed by a processor, analyzes a numerical telemetry signal that comprises state metric data gathered by the monitor, said numerical chart feature detector configured to use a threshold and a prescribed sustain period to detect the presence of a feature based on the numerical telemetry signal, the feature representing a condition or behavior of the software application that is known to be associated with a potential problem of the deployment of the software application; wherein the threshold comprises a zero-confidence level representing substantially zero confidence in the existence of a feature, a trigger-confidence level representing a prescribed intermediate confidence in the existence of a feature, and a full-confidence level representing a substantially maximum confidence in the existence of a feature; wherein the numerical chart feature detector is configured to detect a feature when the telemetry signal exceeds the trigger-confidence level and remains in excess of the zero-confidence level subsequently throughout the prescribed sustain period immediately after exceeding the trigger-confidence level; wherein an occurrence of the feature is represented by at least a name, confidence value, and time of occurrence. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19)
-
-
20. A method of monitoring and analyzing a deployment of a software application, comprising:
-
receiving a numerical telemetry signal that comprises time-variant state metric data collected by a monitor that runs on a computer of said deployment; and analyzing the numerical telemetry signal to determine whether a feature exists, the feature representing a condition or behavior of the software application that is known to be associated with a potential problem of the deployment; wherein analyzing the numerical telemetry signal comprises detecting the feature when the numerical telemetry signal exceeds a trigger-confidence level and then remains in excess of a zero-confidence level throughout a prescribed sustain period, said zero-confidence level corresponding to a substantially zero confidence in the existence of the feature, said trigger-confidence level corresponding to an intermediate confidence that is between the zero-confidence level and a full-confidence level; said method performed by a computer system. - View Dependent Claims (21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33)
-
-
34. A non-transitory computer-readable medium that comprises executable code that, when executed by a computer system, causes the computer system to analyze a deployment of a software application by a process that comprises:
-
receiving a numerical telemetry signal that comprises time-variant state metric data collected by a monitor that runs on a computer of said deployment; and analyzing the numerical telemetry signal to assess whether a feature exists, the feature representing a condition or behavior of the software application that is associated with a potential deployment problem; wherein analyzing the numerical telemetry signal comprises detecting the feature when the numerical telemetry signal exceeds a trigger-confidence level and then remains in excess of a zero-confidence level throughout a prescribed sustain period, said zero-confidence level corresponding to a substantially zero confidence in the existence of the feature, said trigger-confidence level corresponding to an intermediate confidence that is between the zero-confidence level and a full-confidence level. - View Dependent Claims (35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46)
-
Specification