clarify docs

sdebruyn · Mar 31, 2024 · a698846 · a698846
1 parent fa64ed7
commit a698846
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 1 deletion.
diff --git a/docs/usage/continuous-aggregates.md b/docs/usage/continuous-aggregates.md
@@ -8,6 +8,9 @@ Continuous aggregates are the reason that this adapter exists. With this adapter
 !!! tip "Materialized views"
     dbt-postgres 1.6 added support for [materialized views](https://docs.getdbt.com/docs/build/materializations#materialized-view). This feature is **also still available** in this adapter. The main difference between materialized views and continuous aggregates is that continuous aggregates are automatically refreshed (based on a policy) by TimescaleDB, while materialized views are refreshed manually or when you run `dbt run`.
 
+!!! warning "Replaced on each run for now"
+    The current implementation replaces continuous aggregates on each run. This is a limitation of the current implementation and will be improved in the future.
+
 ## Usage
 
 To use continuous aggregates, you need to set the `materialized` config to `continuous_aggregate`.

diff --git a/docs/usage/virtual-hypertables.md b/docs/usage/virtual-hypertables.md
@@ -7,9 +7,12 @@ Hypertables by themselves don't make a lot of sense in dbt as you'd create them
 !!! info
     Consult the [Timescale docs](https://docs.timescale.com/use-timescale/latest/hypertables/about-hypertables/) for more information regarding hypertables.
 
-!!! danger "Existing configurations"
+!!! warning "Existing configurations"
     As soon as you start to manage a hypertable as a virtual hypertable with dbt-timescaledb, dbt will replace existing configurations on every run. This includes the retention policy, compression, and other settings. If you have existing configurations in place, you have to make sure to adjust the dbt configuration accordingly.
 
+!!! warning "Compression and indexes"
+    Compression policies and indexes are not properly implemented yet. You can add them to the configuration, but they will stack up on every run. This is a known issue and will be fixed in the future.
+
 ## Usage
 
 The hypertable has to pre-exist in your database. If the hypertable is not present, dbt will throw an error. Optionally, you can specify dbt's built-in `schema` parameter to reference a hypertable in a different schema.