Method and apparatus for dynamically adjusting the resolution of telemetry signals

• US 7,577,542 B2
• Filed: 04/11/2007
• Issued: 08/18/2009
• Est. Priority Date: 04/11/2007
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Gross, Kenny C., Vacar, Dan, McElfresh, David K.
• Primary Examiner(s): Wachsman; Hal D
• Application Number:
  US11/786,684
• Publication Number:
  US 20080252481A1
• Time in Patent Office: 860 Days
• Field of Search: 702/124, 702/58, 702/59, 702/185, 702/186
• US Class Current: 702/124
• CPC Class Codes: H04Q 9/00 Arrangements in telecontrol or telemetry systems for selectively calling a substation from a main station, in which substation desired apparatus is selected for applying a control signal thereto or for obtaining measured values therefrom

Abstract

One embodiment of the present invention provides a system that dynamically adjusts data resolution during proactive-fault-monitoring in a computer system. During operation, the system temporarily stores high-resolution data for a telemetry signal from the computer system in a buffer. The system then generates low-resolution data for the telemetry signal from the high-resolution data. Next, the system monitors the low-resolution data, and while doing so, determines if an anomaly exists in the low-resolution data. If an anomaly exists in the low-resolution data, the system records the high-resolution data from the buffer on a storage device.

Claims

1. A method for dynamically adjusting data resolution during proactive-fault-monitoring in a computer system, comprising:

while performing the proactive-fault-monitoring, temporarily storing high-resolution data for a telemetry signal from the computer system in a buffer;

generating low-resolution data for the telemetry signal from the high resolution data, which involves downsampling the high resolution data;

monitoring the low-resolution data, and while doing so determining if an anomaly exists in the low-resolution data using a fault-detection technique; and

if an anomaly exists in the low-resolution data, recording the high resolution data from the buffer on a storage device.

2. The method of claim 1, wherein generating the low-resolution data from the high-resolution data involves:

computing a sequence of ensemble averages from the high-resolution data, wherein each ensemble average is the average of the high-resolution data within a corresponding predetermined time interval; and

forming the low-resolution data from the sequence of ensemble averages.

3. The method of claim 1, wherein determining if an anomaly exists in the low-resolution data involves:

applying a sequential probability ratio test (SPRT) to the low-resolution data and a time derivative of the low-resolution data; and

determining that an anomaly exists if the SPRT generates an alarm.

4. The method of claim 1, wherein while an anomaly continues to exist, the method further comprises:

storing the high-resolution data in the buffer while recording the high-resolution data on the storage device;

generating the low-resolution data from the high-resolution data, which involves downsampling the high resolution data; and

monitoring the low-resolution data.

5. The method of claim 1, wherein if the telemetry signal returns to normal, the method further comprises:

recording time information associated with the anomaly; and

stopping recording of the high-resolution data on the storage device.

6. The method of claim 1, wherein the method further comprises performing a root-cause-analysis operation on the high-resolution data recorded on the storage device.

7. A method for dynamically adjusting data resolution during proactive-fault-monitoring of a computer system, comprising:

while performing the proactive-fault-monitoring, sampling a telemetry signal from the computer system at a predetermined sampling rate;

analyzing the sampled telemetry signal to detect and/or predict anomalous activity in the telemetry signal using a fault-detection technique; and

if anomalous activity is detected and/or predicted,

sampling the telemetry signal at an increased sampling rate; and

analyzing the telemetry signal at the increased sampling rate using a fault-detection technique,

wherein dynamically adjusting data resolution improves fault-detection sensitivity and false-alarm avoidance, without requiring system resources to continuously process high-resolution data.

8. The method of claim 7, wherein the method further comprises dynamically adjusting the sampling rate based on the detected and/or predicted anomalous activity of the telemetry signal.

9. The method of claim 7, wherein if the anomalous activity of the telemetry signal returns to normal, the method further comprises restoring the sampling rate to the predetermined sampling rate.

10. A method for dynamically adjusting data resolution during proactive-fault-monitoring of a computer system, comprising:

while performing the proactive-fault-monitoring, sampling a telemetry signal using a predetermined number of bits per sample;

analyzing the sampled telemetry signal to detect and/or predict anomalous activity in the telemetry signal using a fault-detection technique; and

if anomalous activity is detected and/or predicted,

sampling the telemetry signal using an increased number of bits per sample; and

analyzing the telemetry signal at the increased number of bits per sample using a fault-detection technique,

wherein dynamically adjusting data resolution improves fault-detection sensitivity and false-alarm avoidance, without requiring system resources to continuously process high-resolution data.

11. The method of claim 10, wherein while sampling the telemetry signal, the method further comprises dynamically adjusting the number of bits per sample based on the detected and/or predicted anomalous activity of the telemetry signal.

12. The method of claim 10, wherein if the anomalous activity of the telemetry signal returns to normal, the method further comprises resuming sampling of the telemetry signal using the predetermined number of bits per sample.

13. An apparatus that dynamically adjusts data resolution during proactive-fault-monitoring in a computer system, comprising:

a storing mechanism configured to store high-resolution data for a telemetry signal from the computer system in a buffer while performing the proactive-fault-monitoring;

a generating mechanism configured to generate low-resolution data for the telemetry signal from the high-resolution data, which involves downsampling the high resolution data;

a monitoring mechanism configured to monitor the low-resolution data, and while doing so determining if an anomaly exists in the low-resolution data using a fault-detection technique; and

a recording mechanism configured to record the high-resolution data from the buffer into a storage device only if an anomaly exists on the low-resolution data.

14. The apparatus of claim 13, wherein the generating mechanism configured to:

compute a sequence of ensemble averages from the high-resolution data, wherein each ensemble average is the average of the high-resolution data within a corresponding predetermined time interval; and

form the low-resolution data from the sequence of ensemble averages.

15. The apparatus of claim 13, wherein the monitoring mechanism is configured to determine if an anomaly exists in the low-resolution data by:

applying a sequential probability ratio test (SPRT) to the low-resolution data and a time derivative of the low-resolution data; and

determining that an anomaly exists if the SPRT generates an alarm.

16. The apparatus of claim 13,

wherein the recording mechanism is configured to continue to store the high-resolution data in the buffer while recording the high-resolution data on the storage device if an anomaly continues to exist;

wherein the generating mechanism is configured to continue to generate the low-resolution data from the high-resolution data if an anomaly continues to exist; and

wherein the monitoring mechanism is configured to continue to monitor the low-resolution data if an anomaly continues to exist.

17. The apparatus of claim 13, wherein the monitoring mechanism is further configured to:

determine if the telemetry signal returns to normal; and

if so, to

record time information associated with the anomaly; and

stop recording of the high-resolution data on the storage device.

18. The apparatus of claim 13, further comprising an analysis mechanism configured to perform a root-cause-analysis on the high-resolution data recorded on the storage device.

19. An apparatus that dynamically adjusts data resolution during proactive-fault-monitoring of a computer system, comprising:

a sampling mechanism configured to sample a telemetry signal from the computer system at a predetermined sampling rate while performing the proactive-fault-monitoring;

an analysis mechanism configured to analyze the sampled telemetry signal to detect and/or predict anomalous activity in the telemetry signal using a fault-detection technique;

wherein the sampling mechanism is configured to sample the telemetry signal at an increased sampling rate if anomalous activity is detected and/or predicted by the analysis mechanism; and

wherein the analysis mechanism is configured to analyze the telemetry signal at the increased sampling rate using a fault-detection technique,

wherein dynamically adjusting data resolution improves fault-detection sensitivity and false-alarm avoidance, without requiring system resources to continuously process high-resolution data.

20. The apparatus of claim 19, further comprising an adjustment mechanism configured to dynamically adjust the sampling rate based on the detected and/or predicted anomalous activity of the telemetry signal.

21. The apparatus of claim 20, wherein the adjustment mechanism is configured to restore the sampling rate to the predetermined sampling rate when the anomalous activity of the telemetry signal returns to normal.