TimescaleDB — Time Series data on Postgres (Done Right)
# TimescaleDB - https://github.com/timescale/timescaledb - is implemented as an extension on #Postgres, exposing your #TimeSeries data as regular tables, while under the hood, these are actually split out into multiple tables (chunks). Each “chunk” is the time-series data partitioned by a time interval, and an optional partition key (location, user ID, device ID, etc) The obvious advantage — dropping data more than a week old is waaaaay simpler (just drop old chunks!) You also get Higher ingest rates : you’re only writing to — and indexing! — the latest chunks, which means you don’t have to worry about memory/disk issues, indexing rates, vaccuming, etc. etc. Query performance : as good as postgres, and better for time-series stuff (like “merge-append”) Time-oriented features : “time-bucketing” with arbitrary intervals (“ give me the sensor data in groups of 3 hours and 17 minutes ”), last/first aggregates (“ give me the last sensor data recorded in e...