Automated and adaptive threshold setting

• US 8,086,708 B2
• Filed: 09/28/2005
• Issued: 12/27/2011
• Est. Priority Date: 06/07/2005
• Status: Active Grant

• Current Assignee: ServiceNow Incorporated
• Original Assignee: International Business Machines Corporation
• Inventors: Breitgand, David, Henis, Ealan, Shehory, Onn
• Primary Examiner(s): Joo, Joshua
• Application Number:
  US11/237,295
• Publication Number:
  US 20060276995A1
• Time in Patent Office: 2,281 Days
• Field of Search: 700/108, 375/229, 709223-224
• US Class Current: 709/223
• CPC Class Codes: H04L 41/00 Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
• CPC Class Codes: H04L 41/147 for predicting network behaviour
• CPC Class Codes: H04L 41/149 for prediction of maintenance
• CPC Class Codes: H04L 41/5003 Managing SLA; Interaction between SLA and QoS
• CPC Class Codes: H04L 41/5009 Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
• CPC Class Codes: H04L 43/08 Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
• CPC Class Codes: H04L 43/16 Threshold monitoring

Abstract

A method for managing a computer system includes monitoring first violations of a service level objective (SLO) of a service running on the computer system so as to determine a first statistical behavior of the first violations. Second violations of a component performance threshold of a component of the computer system are monitored so as to determine a second statistical behavior of the second violations. A model that predicts the second statistical behavior based on the first statistical behavior is produced. The component performance threshold is automatically adjusted responsively to the model, so as to improve a prediction of the first violations by the second violations.

Claims

1. A method, comprising:

in a computer system that includes at least one component having a performance threshold, transmitting alerts from the computer system to a monitoring system when a performance measure of the at least one component crosses the performance threshold;

detecting, using the monitoring system, violations of a Service Level Objective (SLO) defined for a service running on the computer system;

based on the alerts transmitted to the monitoring system when the performance threshold is crossed, producing a model that predicts future violations of the SLO responsively to the alerts and has a prediction reliability; and

automatically adjusting the performance threshold of the component responsively to the model,

wherein producing the model comprises fitting a first sequence comprising historical values of the violations of the SLO and a second sequence comprising historical values of the performance threshold, and wherein automatically adjusting the performance threshold of the component comprises calculating an updated threshold value based on the fitted sequences and a pre-specified prediction reliability.

2. The method according to claim 1, wherein fitting the first and second sequences comprises applying at least one of a logistic regression process and a polynomial fitting process.

3. The method according to claim 1, wherein producing the model comprises estimating a fitting quality responsively to the model and the first and second sequences, and wherein automatically adjusting the threshold comprises determining whether to update the threshold responsively to the fitting quality.

4. The method according to claim 1, wherein fitting the first and second sequences comprises inserting dummy data points into the first and second sequences, so as to reduce a bias in the calculated updated threshold value.

5. The method according to claim 1, wherein fitting the first and second sequences comprises using only a part of the historical values in the first and second sequences corresponding to at least one of recent events, periodic events and events in which a performance metric value is within a predetermined interval.

6. The method according to claim 1, wherein fitting the first and second sequences comprises applying weights to at least some of the historical values corresponding to at least one of recent events, periodic events, rare events and events in which a performance metric value is within a predetermined interval.

7. Apparatus for managing a computer system that includes at least one component having a performance threshold, the apparatus comprising:

an interface, which is coupled to receive from the computer system alerts when the performance measure of the at least one component crosses the performance threshold, and to further receive from the computer system detected violations of a Service Level Objective (SLO) defined for a service running on the computer system; and

a processor, which is arranged to produce, based on the alerts received from the computer system when the performance threshold is crossed, a model that predicts future violations of the SLO responsively to the alerts and has a prediction reliability, and to automatically adjust the performance threshold of the component responsively to the model,

wherein the processor is arranged to fit a first sequence comprising historical values of the violations of the SLO and a second sequence comprising historical values of the performance threshold in order to produce the model, and to calculate an updated threshold value based on the fitted sequences and a pre-specified prediction reliability.

8. The apparatus according to claim 7, wherein the processor is arranged to apply at least one of a logistic regression process and a polynomial fitting process in order to fit the first and second sequences.

9. The apparatus according to claim 7, wherein the processor is arranged to estimate a fitting quality responsively to the model and the first and second sequences, and to determine whether to update the threshold responsively to the fitting quality.

10. The apparatus according to claim 7, wherein the processor is arranged to insert dummy data points into the first and second sequences when producing the model, so as to reduce a bias in the calculated updated threshold value.

11. The apparatus according to claim 7, wherein the processor is arranged to produce the model using only a part of the historical values in the first and second sequences corresponding to at least one of recent events, periodic events and events in which a performance metric value is within a predetermined interval.

12. The apparatus according to claim 7, wherein the processor is arranged to apply weights to at least some of the historical values corresponding to at least one of recent events, periodic events, rare events and events in which a performance metric value is within a predetermined interval.

13. A method, comprising:

in a computer system that includes at least one component having a performance threshold, transmitting alerts from the computer system to a monitoring system when a performance measure of the at least one component crosses the performance threshold;

detecting, using the monitoring system, violations of a Service Level Objective (SLO) defined for a service running on the computer system;

based on the alerts transmitted to the monitoring system when the performance threshold is crossed, producing a model that predicts future violations of the SLO responsively to the alerts and has a prediction reliability; and

automatically adjusting the performance threshold of the component responsively to the model,

wherein producing the model comprises assigning a first counter to count a first number of occurrences in which the performance threshold is crossed, a second counter to count a second number of the occurrences in which the performance threshold is not crossed, a third counter to count a third number of the occurrences in which the SLO is violated, and a fourth counter to count a fourth number of the occurrences in which the SLO is not violated, and wherein adjusting the performance threshold comprises setting the performance threshold responsively to the first, second, third and fourth counters.

14. The method according to claim 13, wherein the computer system comprises a Storage Area Network (SAN).

15. The method according to claim 13, wherein producing the model comprises estimating false-positive and false-negative rates of the alerts with respect to the violations of the SLO.

16. The method according to claim 15, wherein automatically adjusting the threshold comprises causing the estimated false-positive and false-negative rates to converge to predetermined target values.

17. The method according to claim 15, wherein automatically adjusting the threshold comprises searching for a threshold value that minimizes the estimated false-positive and false-negative rates.

18. Apparatus for managing a computer system that includes at least one component having a performance threshold, the apparatus comprising:

an interface, which is coupled to receive from the computer system alerts when the performance measure of the at least one component crosses the performance threshold, and to further receive from the computer system detected violations of a Service Level Objective (SLO) defined for a service running on the computer system; and

a processor, which is arranged to produce, based on the alerts received from the computer system when the performance threshold is crossed, a model that predicts future violations of the SLO responsively to the alerts and has a prediction reliability, and to automatically adjust the performance threshold of the component responsively to the model,

wherein the processor is arranged to produce the model by assigning a first counter to count a first number of occurrences in which the performance threshold is crossed, a second counter to count a second number of the occurrences in which the performance threshold is not crossed, a third counter to count a third number of the occurrences in which the SLO is violated, and a fourth counter to count a fourth number of the occurrences in which the SLO is not violated, and to adjust the performance threshold responsively to the first, second, third and fourth counters.

19. The apparatus according to claim 18, wherein the computer system comprises a Storage Area Network (SAN).

20. The apparatus according to claim 18, wherein the processor is arranged to estimate false-positive and false-negative rates of the alerts with respect to the violations of the SLO.

21. The apparatus according to claim 20, wherein the processor is arranged to adjust the threshold so as to cause the estimated false-positive and false-negative rates to converge to predetermined target values.

22. The apparatus according to claim 20, wherein the processor is arranged to search for a threshold value that minimizes the estimated false-positive and false-negative rates.

23. A method for performing an interactive analysis of a computer system, which includes at least one component having a performance threshold, to devise an information technology solution applicable to the computer system, the method comprising:

transmitting alerts from the computer system to a monitoring system when a performance measure of the at least one component crosses the performance threshold;

detecting, using the monitoring system, violations of a Service Level Objective (SLO) defined for a service running on the computer system;

based on the alerts transmitted to the monitoring system when the performance threshold is crossed, producing a model that predicts future violations of the SLO responsively to the alerts and has a prediction reliability; and

automatically adjusting the performance threshold of the component responsively to the model,

wherein producing the model comprises fitting a first sequence comprising historical values of the violations of the SLO and a second sequence comprising historical values of the performance threshold, and wherein automatically adjusting the performance threshold of the component comprises calculating an updated threshold value based on the fitted sequences and a pre-specified prediction reliability.

24. The method according to claim 23, wherein the computer system comprises a Storage Area Network (SAN).

25. The method according to claim 23, wherein producing the model comprises estimating false-positive and false-negative rates of the alerts with respect to the violations of the SLO, and wherein automatically adjusting the performance threshold comprises causing the estimated false-positive and false-negative rates to converge to predetermined target values.

26. The method according to claim 25, wherein automatically adjusting the threshold comprises searching for a threshold value that minimizes the estimated false-positive and false-negative rates.