Analytics-correlation-runbooks pipeline for surfacing actionable insights

A non-transitory computer-readable medium comprising stored instructions, which when executed by a processor, cause the processor to:

• store, in a monitoring database, measurement data received from a set of sensors in a computing infrastructure;
• determine, by an anomaly engine, an indicator when at least a first measurement of a monitored metric is outside a range of expected values, wherein the anomaly engine creates a time series profile for the monitored metric based on historical measurement of the monitored metric, and wherein the range of expected values is generated based on the time series profile for the monitored metric;
• determine, by a correlation engine, a detection notification when the indicator matches a detection pattern;
• execute, by a runbook engine, a runbook based on the detection notification, wherein at least one command in the runbook queries the monitoring database for a second measurement which is different from the first measurement;
• and generate an actionable insight based on a result produced by the runbook.

A system, comprising:

• a set of memories storing instructions;
• and
• a set of processors, coupled with the set of memories, and to execute the instructions, the instructions when executed causing the set of processors to:
• store, in a monitoring database, measurement data received from a set of sensors in a computing infrastructure;
• determine, by an anomaly engine, an indicator when at least a first measurement of a monitored metric is outside a range of expected values, wherein the anomaly engine creates a time series profile for the monitored metric based on historical measurement of the monitored metric, and wherein the range of expected values is generated based on the time series profile for the monitored metric;
• determine, by a correlation engine, a detection notification when the indicator matches a detection pattern;
• execute, by a runbook engine, a runbook based on the detection notification, wherein a first command in the runbook queries the monitoring database for a second measurement which is different from the first measurement, and wherein a second command in the runbook queries an external database which is different from the monitoring database;
• and
• generate an actionable insight based on a result produced by the runbook.