Design for Continuous Query

Continuous Query is a very useful feature for users to query time series data automatically and periodically on realtime data and store query results in a specified measurement according to the time interval and time range.

For example, using the continuous query described below, the IoTDB downsamples the CPU load metric, and keeps a single average datapoint per 15 minutes grouped by server region:

CREATE CONTINUOUS QUERY "downsampled_cpu_load"
ON database_name
BEGIN
  SELECT mean(value) as value,
  INTO "downsampled.cpu_load"
  FROM cpu_load
  GROUP BY time(15m), region
END

Syntax

Basic syntax