[Feature]: CREATE OR REPLACE continuous aggregate

[luukvnes]

What problem does the new feature solve?

Hi,

I have continous aggregates that holds data from a hypertable where that data is already dropped. Whenever i want to change the continous aggregate in a meaningful way (e.g. alter, drop or instert a column) I have a number of unpleasant solutions.

• Hold a duplicate cagg until the data is past the retention policy of the cagg (takes up a lot of space)
• Create a temporary hypertable that copies data from the original cagg before basing the the new cagg on that hypertable
• Keep the old cagg and create an additional cagg with only the changed columns (becomes very messy)

Therefore being able to edit a the cagg without dropping data would help a lot.

What does the feature do?

Allow a CREATE OR REPLACE for a continuous aggregate.

Implementation challenges

No response

[salisbury-espinosa]

e.g. alter, drop or instert a column

Allow a CREATE OR REPLACE for a continuous aggregate.

Timescale continuous aggregates use PostgreSQL materialized views to continuously and incrementally refresh a query in the background

that's why you can't:

https://www.postgresql.org/docs/current/sql-altermaterializedview.html