Automated selection of performance monitors

1. A computer implemented method comprising:

deploying a plurality of application programming interfaces (APIs) in a cloud computing infrastructure;

extracting, by a processor, statistical data from an API call log, the statistical data being associated with an invocation of an API by an application;

calculating, by the processor, a performance value associated with the invocation of the API using the statistical data;

generating, by the processor, an aggregate dataset that combines the performance value associated with the invocation of the API with performance values associated with respective previous invocations of the API;

calculating, by the processor, metric values for respective time intervals of a time period, wherein the metric values are associated with performance values during the respective time intervals;

calculating, by the processor, mean and standard deviation values of the metric values for the time period;

selecting, by the processor, the API as a candidate API, the API being selected from among the plurality of APIs based at least in part on criteria involving performance values of the APIs, wherein the criteria comprise (i) an arrival rate of the candidate API is within a band of highest arrival rates for the plurality of APIs, (ii) a standard deviation of a metric for the candidate API is not greater than a mean of the metric for the API, and (iii) a ratio ρ between a mean arrival rate λ for the candidate API and a mean response time, μ for the candidate API times a number of replicas c for the candidate API according to Expression p=λ/c·μ is less than 1;

detecting, by the processor, a Customer Impacting Event (CIE) by applying a machine learning algorithm using monitored values associated with the candidate API during a time frame defined by a rolling window; and

responsive to the detecting of the CIE, automatically initiating, by the processor, a selected alert from among a plurality of alert options based at least in part on the monitored values associated with the CIE.

2. The computer implemented method of claim 1 , wherein the statistical data includes response time data and arrival rate data.

3. The computer implemented method of claim 1 , wherein the extracting further comprises extracting statistical data from the API call log associated with invocation of a plurality of APIs, the plurality of APIs including said API.

4. The computer implemented method of claim 1 , wherein the calculating of a performance value comprises calculating a response time.

5. The computer implemented method of claim 1 , wherein the calculating of a performance value further comprises calculating an arrival rate.

6. The computer implemented method of claim 1 , wherein a type of metric used for calculating the metric values is selected from the group consisting of a median value, a 95th percentile value, and a 99th percentile value.

7. The computer implemented method of claim 1 , wherein the time intervals include a plurality of equal time intervals.

8. The computer implemented method of claim 1 , wherein the machine learning algorithm provides an indication of monitored values that are outlier values indicative of the CIE.

9. The computer implemented method of claim 1 , wherein the monitored values comprise response time values and arrival rate values.

10. The computer implemented method of claim 1 , wherein the plurality of alert options include an outage alert and a performance alert.

11. A computer program product for automated selection of performance and reliability monitors, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform operations comprising:

deploying a plurality of application programming interfaces (APIs) in a cloud computing infrastructure;

extracting, by a processor, statistical data from an API call log, the statistical data being associated with an invocation of an API by an application;

calculating, by the processor, a performance value associated with the invocation of the API using the statistical data;

generating, by the processor, an aggregate dataset that combines the performance value associated with the invocation of the API with performance values associated with respective previous invocations of the API;

calculating, by the processor, metric values for respective time intervals of a time period, wherein the metric values are associated with performance values during the respective time intervals;

calculating, by the processor, mean and standard deviation values of the metric values for the time period;

selecting, by the processor, the API as a candidate API, the API being selected from among the plurality of APIs based at least in part on criteria involving performance values of the APIs, wherein the criteria comprise (i) an arrival rate of the candidate API is within a band of highest arrival rates for the plurality of APIs, (ii) a standard deviation of a metric for the candidate API is not greater than a mean of the metric for the API, and (iii) a ratio ρ between a mean arrival rate λ for the candidate API and a mean response time μ for the candidate API times a number of replicas c for the candidate API according to Expression p=λ/c·μ is less than 1;

detecting, by the processor, a CIE by applying a machine learning algorithm using monitored values associated with the candidate API during a time frame defined by a rolling window; and

responsive to the detecting of the CIE, automatically initiating, by the processor, a selected alert from among a plurality of alert options based at least in part on the monitored values associated with the CIE.

12. The computer program product of claim 11 , wherein the stored program instructions are stored in a computer readable storage device in a data processing system, and wherein the stored program instructions are transferred over a network from a remote data processing system.

13. The computer program product of claim 11 , wherein the stored program instructions are stored in a computer readable storage device in a server data processing system, and wherein the stored program instructions are downloaded over a network to a remote data processing system for use in a computer readable storage device associated with the remote data processing system, further comprising:

metering use of the computer usable code associated with the request; and

generating an invoice based on the metered use.

14. The computer program product of claim 11 , wherein the statistical data includes response time data and arrival rate data.

15. The computer program product of claim 11 , wherein a type of metric used for calculating the metric values is selected from the group consisting of a median value, a 95th percentile value, and a 99th percentile value.

16. The computer program product of claim 11 , wherein the plurality of alert options include an outage alert and a performance alert.

17. A computer system comprising a processor and one or more computer readable storage media, and program instructions collectively stored on the one or more computer readable storage media, the program instructions executable by the processor to cause the processor to perform operations comprising:

deploying a plurality of application programming interfaces (APIs) in a cloud computing infrastructure;

extracting, by a processor, statistical data from an API call log, the statistical data being associated with an invocation of an API by an application;

calculating, by the processor, a performance value associated with the invocation of the API using the statistical data;

generating, by the processor, an aggregate dataset that combines the performance value associated with the invocation of the API with performance values associated with respective previous invocations of the API;

calculating, by the processor, metric values for respective time intervals of a time period, wherein the metric values are associated with performance values during the respective time intervals;

calculating, by the processor, mean and standard deviation values of the metric values for the time period;

selecting, by the processor, the API as a candidate API, the API being selected from among the plurality of APIs based at least in part on criteria involving performance values of the APIs, wherein the criteria comprise (i) an arrival rate of the candidate API is within a band of highest arrival rates for the plurality of APIs, (ii) a standard deviation of a metric for the candidate API is not greater than a mean of the metric for the API, and (iii) a ratio ρ between a mean arrival rate λ for the candidate API and a mean response time μ for the candidate API times a number of replicas c for the candidate API according to Expression p=λ/c·μ is less than 1;

detecting, by the processor, a CIE by applying a machine learning algorithm using monitored values associated with the candidate API during a time frame defined by a rolling window; and

responsive to the detecting of the CIE, automatically initiating, by the processor, a selected alert from among a plurality of alert options based at least in part on the monitored values associated with the CIE.

18. The computer system of claim 17 , wherein the statistical data includes response time data and arrival rate data.

19. The computer system of claim 17 , wherein a type of metric used for calculating the metric values is selected from the group consisting of a median value, a 95th percentile value, and a 99th percentile value.

20. The computer system of claim 17 , wherein the plurality of alert options include an outage alert and a performance alert.