Time Series Data

Time series is data measured over a time interval in successive order. Each data sample comes with an attached timestamp. Time series examples are CPU utilization, memory utilization, disk read and write, or temperature.

Time series data can be pushed into ATSD using network api and data api, inserted as CSV files, or pulled from external data sources using Axibase Collector.

Date,Entity,Metric,Value
22-11-13 09:25:00,host01,load_avg_1,0.1
22-11-13 09:30:00,host01,load_avg_1,0.1
22-11-13 09:35:00,host01,load_avg_1,0.2
22-11-13 09:40:00,host01,load_avg_1,0.4
22-11-13 09:45:00,host01,load_avg_1,0.7

When storing time series data, new metrics are initialized with float data type. The data type can be then modified to be decimal, double, float, long, integer, and short.

The primary key for a time series array is metric name, entity name, and optional series tags. The list of metrics can be found on the Metrics page.

Time Series Data is the primary component of all Portals and Widgets. This data type is graphed below.