Time series as a native data structure

Time Series automatically executes downsampling and retention rules with double-delta compression to space-efficiently store large time-series datasets.

On the left are the raw data samples. The high point cardinality obscures the overall trend and requires more storage. On the right is a downsampled representation of the same data. Time Series can automatically perform downsampling by aggregating many points over time, reducing both noise in historical data and the storage requirements.

Powered by labelling and search techniques, implement multiple range queries and aggregations across several time-series objects for real-time analysis. Use counter operations such as increment and decrement on the last value for telemetry apps.