Combining multiple signals to determine global system state

• US 9,276,826 B1
• Filed: 03/22/2013
• Issued: 03/01/2016
• Est. Priority Date: 03/22/2013
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Dickson, Caskey
• Primary Examiner(s): Gilles, Jude Jean
• Application Number:
  US13/849,163
• Time in Patent Office: 1,075 Days
• Field of Search: 709/217, 709/224, 709/223
• US Class Current: 1/1
• CPC Class Codes: H04L 41/145 involving simulating, designing, planning or modelling of a network
• 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 43/12 Network monitoring probes
• CPC Class Codes: H04L 49/00 Packet switching elements

Abstract

Methods and systems are provided for monitoring the health of a service by using a comprehensive collection of analytical values derived from data obtained from a plurality of sensors. Many simultaneous data points are gathered by sensors monitoring a service, which then send the raw data to a collection of analyzers (e.g., servers) running in one or more centralized data centers for examination. Collected data is analyzed to generate various metrics (e.g., analytical values) that can be combined with other data and analysis results to provide indications about the overall health of the service.

Claims

1. A monitoring system comprising:

one or more processing devices; and

one or more storage devices storing instructions that, when executed by the one or more processing devices, cause the one or more processing devices to perform;

receiving sensor data obtained for a network service from a plurality of sensors at one or more locations of the network service;

combining the sensor data for the network service received from the plurality of sensors;

analyzing the combined sensor data for the network service;

generating one or more metrics for evaluating performance of the network service, the one or more metrics being based on the analyzed combined sensor data; and

storing the one or more metrics in a data repository.

2. The system of claim 1, wherein the one or more processing devices are further caused to perform retrieving, from at least one database, historical sensor data for the network service.

3. The system of claim 2, wherein the one or more processing devices are further caused to perform:

combining the sensor data for the network service received from the plurality of sensors with the historical sensor data for the network service retrieved from the at least one database; and

analyzing the combined sensor data and historical sensor data to generate one or more metrics for evaluating performance of the network service.

4. The system of claim 1, wherein the one or more processing devices are further caused to perform retrieving, from at least one database, administrative data for the network service, the administrative data indicating an expected status of the network service.

5. The system of claim 4, wherein the one or more processing devices are further caused to perform:

combining the sensor data for the network service received from the plurality of sensors with the administrative data for the network service retrieved from the at least one database; and

analyzing the combined sensor data and administrative data to generate one or more metrics for evaluating performance of the network service.

6. The system of claim 1, further comprising an administrative database configured to store data about an operational status of the network service.

7. The system of claim 1, further comprising a health history database configured to store data about historical observations for the network service.

8. The system of claim 1, further comprising an alerting component configured to generate an alert signal based on the one or more metrics for evaluating performance of the network service.

9. The system of claim 8, wherein the one or more processing devices are further caused to perform sending a notification of the one or more metrics being stored in the data repository to the alerting component.

10. The system of claim 1, wherein the sensor data includes processed and/or unprocessed sensor data for the network service.

11. The system of claim 1, wherein the one or more metrics include one or more of operating efficiency of the network service, overloaded components of the network service, and failing components of the network service.

12. A computer-implemented method for monitoring a network service, the method comprising:

receiving sensor data obtained for the network service from a plurality of sensors at one or more locations of the network service;

combining the sensor data for the network service received from the plurality of sensors;

analyzing the combined sensor data for the network service;

generating one or more metrics for evaluating performance of the network service, the one or more metrics being based on the analyzed combined sensor data; and

storing the one or more metrics in a data repository.

13. The method of claim 12, further comprising:

obtaining, from a database, historical sensor data stored for the network service;

combining the sensor data received from each of the sensors with the historical sensor data obtained from the database; and

generating one or more metrics for evaluating performance of the network service based on the combined sensor data and historical sensor data.

14. The method of claim 12, further comprising;

obtaining, from a database, administrative data for the network service, the administrative data indicating an expected status of the network service;

combining the sensor data received from each of the sensors with the administrative data obtained from the database; and

generating one or more metrics for evaluating performance of the network service based on the combined sensor data and administrative data.

15. The method of claim 12, further comprising:

obtaining, from one or more databases, historical sensor data and administrative data stored for the network service, the historical sensor data including historical observations for the network service and the administrative data indicating an expected status of the network service;

generating one or more metrics for evaluating performance of the network service based on the sensor data received from the sensors, the historical sensor data, and the administrative data; and

storing the computed one or more metrics, the sensor data, the historical sensor data, and the administrative data in a data repository.

16. The method of claim 12, further comprising generating a notification that the one or more metrics are stored in the data repository.

17. A computer-implemented method comprising:

receiving, at a server, a service health request for a network service;

determining that the service health request is valid based on the service health request being consistent with the network service and resources being available to fulfill the request;

collecting sensor data for the network service;

storing the service health request and the collected sensor data in a data repository; and

generating a notification indicating that the service health request and the collected sensor data for the network service are stored in the data repository.

18. The method of claim 17, wherein the service health request is a request for the network service to be monitored.

19. The method of claim 17, wherein the service health request is a request to verify a connection with the network service.

20. The method of claim 17, further comprising:

determining that the request is consistent with the network service; and

determining that resources are available to fulfill the request.

21. The method of claim 17, wherein collecting sensor data for the network service includes:

sending a data collection request to sensors at one or more locations of the network service; and

receiving, in response to the data collection request, sensor data from each of the sensors.

22. The method of claim 17, further comprising:

obtaining, from one or more databases, data about historical observations made for the network service and data about expected status of the network service; and

storing the data about historical observations and the data about expected status in the data repository with the service health request and collected sensor data.

23. A computer-implemented method comprising:

receiving, at a server, a notification that a service health request has been made for a network service;

retrieving, from a data repository, the service health request and data collected for the network service;

determining, based on the service health request, one or more analyses to be performed;

performing the one or more analyses using the data collected for the network service to compute one or more metrics for evaluating performance of the network service; and

storing the one or more computed metrics with the service health request in the data repository.

24. The method of claim 23, wherein the data collected for the network service includes one or more of sensor data obtained by sensors at one or more locations of the network service, data about historical observations made for the network service, and administrative data for the network service, the administrative data indicating an expected status of the network service.

25. The method of claim 23, further comprising generating a notification indicating that the service health request and the one or more computed metrics for the network service are stored in the data repository.