Telemetry data filtering through sequential analysis

A system for analyzing telemetry data from a computer system, comprising:

• a processor;
• a monitoring mechanism configured to periodically obtain the telemetry data from the computer system;
• a filtering apparatus configured to detect a statistical deviation in the telemetry data using a sequential-analysis technique, wherein the sequential-analysis technique is a sequential probability ratio test;
• and
• an analysis apparatus configured to, in response to the detection:
• identify a subset of the telemetry data associated with the statistical deviation;
• apply a root-cause-analysis technique to the subset of the telemetry data to determine a source of the statistical deviation, wherein the root-cause-analysis technique is a multivariate state estimation technique;
• and
• use the source of the statistical deviation to perform a remedial action for the computer system, which involves correcting a fault in the computer system corresponding to the source of the statistical deviation.

A computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method for analyzing telemetry data from a computer system, the method comprising:

• periodically obtaining the telemetry data from the computer system;
• preprocessing the telemetry data using a sequential-analysis technique;
• detecting a statistical deviation in the telemetry using the sequential analysis technique, wherein the sequential analysis technique is a sequential probability ratio test;
• in response to the detection,
• identifying a subset of the telemetry data associated with the statistical deviation,
• applying a root-cause-analysis technique to the subset of the telemetry data to determine a source of the statistical deviation, wherein the root-cause-analysis technique is a multivariate state estimation technique, and
• using the source of the statistical deviation to perform a remedial action for the computer system, which involves correcting a fault in the computer system corresponding to the source of the statistical deviation.