Data metrics analytics

• US 9,992,090 B2
• Filed: 01/08/2014
• Issued: 06/05/2018
• Est. Priority Date: 01/08/2014
• Status: Active Grant

• Current Assignee: Bank of America Corp.
• Original Assignee: Bank of America Corp.
• Inventors: Nayyar, Ashish K., Dixit, Amit Subhash, Piduru, Trivikram
• Primary Examiner(s): Christensen, Scott B
• Application Number:
  US14/150,252
• Publication Number:
  US 20150195157A1
• Time in Patent Office: 1,609 Days
• CPC Class Codes: G06F 11/00 Error detection; Error correction; Monitoring error detection, correction or monitoring in information storage based on relative movement between record carrier and transducer G11B20/18; monitoring, i.e. supervising the progress of recording or reproducing G11B27/36; in static stores G11C29/00
• CPC Class Codes: G06F 11/3003 Monitoring arrangements specially adapted to the computing system or computing system component being monitored
• CPC Class Codes: G06F 11/32 with visual or acoustical indication of the functioning of the machine
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 11/3452 Performance evaluation by statistical analysis
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 2201/81 Threshold
• CPC Class Codes: G06F 2201/865 Monitoring of software
• CPC Class Codes: G06F 2201/875 Monitoring of systems including the internet
• CPC Class Codes: G06F 2209/508 Monitor
• CPC Class Codes: H04L 41/0681 Configuration of triggering conditions
• CPC Class Codes: H04L 43/0852 Delays

Abstract

Aspects of the present disclosure provide systems and methods directed toward identifying causes of attenuated performance at computing systems. Performance data corresponding to performance metrics may be stored at a data store. Performance metrics may be associated with application servers. Alerts may be received from a performance monitoring agent triggered by performance attenuations at the application servers. A timeframe in which a performance attenuation occurred may be selected and the performance data may be searched in order to identify another performance attenuation that occurred at one of the application servers during or before the selected timeframe. Performance data may be evaluated to identify one of the application servers associated with the additional performance attenuation as a potential source of the performance attenuation that triggered the alert. Information corresponding to the potential source of the performance attenuation that triggered the alert may be presented to a user at a display device.

Claims

1. A computer-implemented method of identifying a cause of attenuated performance at a computing system comprising:

storing, at a data store of a computing device, performance data corresponding to a plurality of performance metrics associated with one or more application servers of a computing system and monitored by a performance monitoring agent;

receiving, by the computing device, an alert from the performance monitoring agent, the alert having been triggered by a first performance attenuation at one of the one or more application servers;

selecting, by the computing device, a timeframe in which the first performance attenuation occurred;

searching, by the computing device, the performance data to identify a second performance attenuation that occurred at one of the one or more application servers during or before the timeframe selected;

evaluating, by the computing device, the performance data to identify one of the one or more application servers at which the second performance attenuation occurred as a potential source of the first performance attenuation that triggered the alert, wherein evaluating the performance data comprises iterating over the plurality of performance metrics and, for each performance metric of the plurality of performance metrics,

determining, based on the performance data, that the performance metric crossed a predetermined performance threshold used to trigger alerts during or before the timeframe selected,

identifying, in response to the determining, the performance metric as an anomalous performance metric, and

selecting one of the one or more application servers associated with the anomalous performance metric as the potential source of the first performance attenuation;

determining, by the computing device, whether at least one of;

a volume of traffic or network latency has increased;

responsive to determining that the at least one of;

the volume of traffic or the network latency has increased, determining whether the at least one of;

the increased volume or traffic or increased network latency corresponds to the first performance attenuation;

responsive to determining that the at least one of;

the increased volume of traffic or the increased network latency corresponds to the first performance attenuation, adjusting, by the computing device, a performance threshold used to trigger alerts associated with one of the plurality of performance metrics; and

presenting, at a display device of the computing device, information corresponding to at least one of the application servers identified as one of the one or more potential sources of the first performance attenuation.

2. The method of claim 1 wherein:

the performance data includes alerts triggered for the one or more of the application servers;

searching the performance data includes identifying an alert triggered for one of the application servers during or before the timeframe selected; and

the application server identified as the potential source of the first performance attenuation is the application server of the one or more application servers for which an alert was triggered during or before the timeframe selected.

3. The method of claim 1 wherein at least some of the performance metrics are associated with one or more applications of the computing system and further comprising:

selecting, by the computing device, a first application of the one or more applications for which to create a dependency map;

identifying, based on the performance data, a first previous performance attenuation associated with the first application and a second previous performance attenuation associated with a second application;

determining that the first previous performance attenuation correlates with the second previous performance attenuation;

configuring the dependency map for the first application to include a relationship between the first application and the second application; and

storing the dependency map at the data store.

4. The method of claim 3 further comprising:

determining that the first previous performance attenuation associated with the first application correlates with the second previous performance attenuation associated with the second application when the first previous performance attenuation occurs within a predetermined time period of the second previous performance attenuation.

5. The method of claim 4 further comprising:

determining whether the first previous performance attenuation associated with the first application occurred before or after the second previous performance attenuation associated with the second application; and

configuring the dependency map to identify the first application as a downstream application or an upstream application relative to the second application responsive to a determination of whether the first previous performance attenuation occurred before or after the second previous performance attenuation.

6. The method of claim 5 further comprising:

receiving, by the computing device, a subsequent alert from the performance monitoring agent triggered by a subsequent performance attenuation at the first application;

identifying, by the computing device and based on the dependency map, one or more downstream applications related to the first application; and

transmitting, by the computing device, to each of the one or more downstream applications a notification indicating the subsequent performance attenuation at the first application.

7. The method of claim 1 wherein:

the second performance attenuation is associated with an application hosted by one of the one or more application servers identified as the potential source of the first performance attenuation that triggered the alert.

9. The method of claim 1, further comprising:

storing, at the data store, a dependency map associated with an application residing at one of the one or more application servers identified as one of the one or more potential sources of the first performance attenuation, the dependency map indicating one or more additional applications downstream of the application; and

providing, by the computing device for each of the one or more applications, a notification of the second performance attenuation.

10. The method of claim 1, wherein:

adjusting the performance threshold comprises automatically adjusting the performance threshold upon starting a peak usage period or an off-peak usage period.

8. A system for identifying a cause of attenuated performance at a computing system comprising:

one or more processors;

a display device;

a data store that stores performance data corresponding to a plurality of performance metrics associated with one or more application servers of a computing system and monitored by a performance monitoring agent;

memory storing instructions that, when executed by one of the one or more processors, cause the system to;

evaluate, automatically in response to detecting a performance attenuation at a first application server of the one or more application servers, the performance data to identify a second application server as one of one or more potential sources of the performance attenuation, wherein evaluating the performance data comprises iterating over the plurality of performance metrics and, for each performance metric,

determine, based on the performance data, that the performance metric crossed a predetermined performance threshold used to trigger alerts during or before a timeframe the performance attenuation occurred,

identify, in response, the performance metric as an anomalous performance metric, and

select one of the one or more application servers associated with the anomalous performance metric as one of the one or more potential sources of the performance attenuation;

determining, by the computing system, whether at least one of;

a volume of traffic or network latency has increased;

responsive to determining that the at least one of;

the volume of traffic or the network latency has increased, determining whether the at least one of;

the increased volume or traffic or increased network latency corresponds to the first performance attenuation;

responsive to determining that the at least one of;

the increased volume of traffic or the increased network latency corresponds to the first performance attenuation, adjusting a performance threshold used to trigger alerts associated with one of the plurality of performance metrics; and

present, a the display device, information corresponding to at least one of the one or more application servers identified as one of the one or more potential sources of the performance attenuation.

11. The system of claim 8 wherein:

the performance attenuation is a first performance attenuation; and

evaluating the performance data further comprises

searching the performance data for a second performance attenuation that correlates with the first performance attenuation, and

identifying one of the one or more application servers that is associated with the second performance attenuation as one of the one or more potential sources of the first performance attenuation.

12. The system of claim 11 wherein:

determining that the second performance attenuation correlates with the first performance attenuation comprises determining that the second performance attenuation occurred within a predetermined time period of the first performance attenuation.

13. The system of claim 8 wherein:

the instructions, when executed by one of the one or more processors, further cause the system to

generate a dependency map for an application residing at one of the one or more application servers based on evaluation of the performance data.

14. The system of claim 13 wherein:

the application is a first application; and

generating the dependency map comprises

(i) establishing a relationship between the first application and a second application in the dependency map responsive to a determination that the first application is associated with a first previous performance attenuation that correlates with a second previous performance attenuation associated with the second application, and

(ii) identifying the first application as a downstream application relative to the second application in the dependency map responsive to a determination that the first previous performance attenuation occurred after the second previous performance attenuation, or identifying the first application as an upstream application relative to the second application in the dependency map responsive to a determination that the first previous performance attenuation occurred before the second previous performance attenuation.

15. The system of claim 14 further comprising:

the instructions, when executed by one of the one or more processors, further cause the system to

receive an alert from the performance monitoring agent indicating a subsequent performance attenuation at the first application,

identify one or more downstream applications related to the first application based on the dependency map, and

transmit for each of the one or more downstream applications a notification indicating the subsequent performance attenuation at the first application.

16. The system of claim 15 further comprising:

the instructions, when executed by one of the one or more processors, further cause the system to;

revert the performance threshold responsive to a determination that the volume of traffic or the network latency has decreased.