Convert continuous aggregates to the columnstore

URL: llms-txt#convert-continuous-aggregates-to-the-columnstore

Contents:

• Enable compression on continuous aggregates
  • Enabling and disabling compression on continuous aggregates
• Compression policies on continuous aggregates

Continuous aggregates are often used to downsample historical data. If the data is only used for analytical queries and never modified, you can compress the aggregate to save on storage.

Old API since TimescaleDB v2.18.0 Replaced by <a href="https://docs.tigerdata.com/use-timescale/latest/continuous-aggregates/compression-on-continuous-aggregates/">Convert continuous aggregates to the columnstore</a>.

Before version 2.18.1, you can't refresh the compressed regions of a continuous aggregate. To avoid conflicts between compression and refresh, make sure you set compress_after to a larger interval than the start_offset of your refresh policy.

Compression on continuous aggregates works similarly to [compression on hypertables][compression]. When compression is enabled and no other options are provided, the segment_by value will be automatically set to the group by columns of the continuous aggregate and the time_bucket column will be used as the order_by column in the compression configuration.