Identifying deviations in data
First Claim
Patent Images
1. A method implemented in operation control information technology environment, to identify metrics that cause a deviation in data, comprising:
- collecting, by a processor, the data for selected metrics stored in a plurality of tables, wherein the data includes operational data fetched from one or more system components comprising servers, network components or storage components;
constructing a metric vector based on the data for the selected metrics,wherein the selected metrics include a percent of memory used by a server, a percent of a computer processing unit (CPU) used by the server, or an input/output utilization of the server monitored over a period of time;
calculating a probability density for the metric vector that indicates a deviation value for the metric vector relative to other metric vectors,wherein the calculating of the probability density for the metric vector includes implementing a Multivariate Gaussian Distribution algorithm; and
identifying an outlier metric from the metric vector that causes the deviation value for the metric vector, wherein the identifying of the outlier metric includes;
selecting a maximum outlier product from the multiplication of (x−
μ
)TΣ
−
1 and (x−
μ
), where x is the metric vector, μ
is a mean distribution vector, and Σ
is a covariance matrix, anddetermining the outlier metric based on the maximum outlier product; and
detecting anomaly associated with the one or more system components based on the outlier metric.
2 Assignments
0 Petitions
Accused Products
Abstract
In an example, metrics that cause a deviation in data may be identified by collecting the data for selected metrics stored in a plurality of tables. A metric vector is constructed based on the data for the selected metrics. A probability density may be calculated for the metric vector that indicates a deviation value for the metric vector relative to other metric vectors. Moreover, an outlier metric from the metric vector that causes the deviation value for the metric vector may be identified.
-
Citations
12 Claims
-
1. A method implemented in operation control information technology environment, to identify metrics that cause a deviation in data, comprising:
-
collecting, by a processor, the data for selected metrics stored in a plurality of tables, wherein the data includes operational data fetched from one or more system components comprising servers, network components or storage components; constructing a metric vector based on the data for the selected metrics, wherein the selected metrics include a percent of memory used by a server, a percent of a computer processing unit (CPU) used by the server, or an input/output utilization of the server monitored over a period of time; calculating a probability density for the metric vector that indicates a deviation value for the metric vector relative to other metric vectors, wherein the calculating of the probability density for the metric vector includes implementing a Multivariate Gaussian Distribution algorithm; and identifying an outlier metric from the metric vector that causes the deviation value for the metric vector, wherein the identifying of the outlier metric includes; selecting a maximum outlier product from the multiplication of (x−
μ
)TΣ
−
1 and (x−
μ
), where x is the metric vector, μ
is a mean distribution vector, and Σ
is a covariance matrix, anddetermining the outlier metric based on the maximum outlier product; and
detecting anomaly associated with the one or more system components based on the outlier metric. - View Dependent Claims (2, 3, 4, 5, 6, 7)
-
-
8. A system for operation control information technology environment, to identify metrics that cause a deviation in data, comprising:
-
a processor and a memory; a data collector engine stored in a memory, executed by the processor, to collect the data for selected metrics stored in a plurality of tables, wherein the data includes operational data fetched from one or more system components comprising servers, network components or storage components; a vector generating engine, stored in the memory, to construct a metric vector based on the data for the selected metrics, wherein the selected metrics include a percent of memory used by a server, a percent of a computer processing unit (CPU) used by the server, or an input/output utilization of the server monitored over a period of time; a probability engine, stored in the memory, to calculate a probability density for the metric vector using a Multivariate Gaussian Distribution algorithm, wherein the probability density indicates a deviation value for the metric vector relative to other metric vectors; and an outlier engine, stored in the memory, to identify an outlier metric from the metric vector that causes the deviation value for the metric vector, wherein the identifying of the outlier metric includes; selecting a maximum outlier product from the multiplication of (x−
μ
)TΣ
−
1 and (x−
μ
), where x is the metric vector, μ
is a mean distribution vector, and Σ
is a covariance matrix, anddetermining the outlier metric based on the maximum outlier product; and
detecting anomaly associated with the one or more system components based on the outlier metric. - View Dependent Claims (9, 10, 11)
-
-
12. A non-transitory computer readable medium including machine readable instructions executable by a processor in operation control information technology environment, to:
-
collect data for selected metrics stored in a plurality of tables, wherein the data includes operational data fetched from one or more system components comprising servers, network components or storage components; construct a metric vector based on the data for the selected metrics, wherein the selected metrics include a percent of memory used by a server, a percent of a computer processing unit (CPU) used by the server, or an input/output utilization of the server monitored over a period of time; calculate a probability density for the metric vector using a Multivariate Gaussian Distribution algorithm, wherein the probability density indicates a deviation value for the metric vector relative to other metric vectors; and identify an outlier metric from the metric vector that causes the deviation value for the metric vector, wherein the identifying of the outlier metric includes; selecting a maximum outlier product from the multiplication of (x−
μ
)TΣ
−
1 and (x−
μ
), where x is the metric vector, μ
is a mean distribution vector, and Σ
is a covariance matrix, anddetermining the outlier metric based on the maximum outlier product; and
detecting anomaly associated with the one or more system components based on the outlier metric.
-
Specification