I’m having a rough time making my column description persist in Redshift tables. I’ve gone through the related documentation:
I’ve tried configuring it at the model level in the config block:
{{ config(
sort=['load_timestamp'],
dist='load_timestamp',
unique_key="",
alias = "persist_docs_trial",
persist_docs={"relation": true, "columns": true}
) }}
and at the dbt_project.yml at different levels
models:
models:
redacted:
+persist_docs:
relation: true
columns: true
stg:
stg_model:
...
Also at the model level:
models:
redacted:
consumption:
redacted:
materialized: incremental
redacted:
+persist_docs:
relation: true
columns: true
alias: persist_docs_trial
description: Staging table from star model
columns:
- name: id_key_fact_redacted
description: Redacted unique identifier
I can confirm that this is possible in my specific environment since it was successfully implemented based on the following logs but I’m being unlucky reproducing it.
[0m15:27:32.300608 [debug] [Thread-1 ]: SQL status: SELECT in 3 seconds
e[0m15:27:32.311996 [debug] [Thread-1 ]: Using redshift connection "model.redacted.persist_docs_trial"
e[0m15:27:32.312795 [debug] [Thread-1 ]: On model.redacted.persist_docs_trial: /* {"app": "dbt", "dbt_version": "1.4.9", "profile_name": "default", "target_name": "dev", "node_id": "model.redacted.persist_docs_trial"} */
comment on column "redacted"."redacted"."persist_docs_trial".id_key_fact_redacted is $dbt_comment_literal_block$Unique identifiert$dbt_comment_literal_block$;
e[0m15:27:32.423812 [debug] [Thread-1 ]: SQL status: COMMENT in 0 seconds
e[0m15:27:32.431695 [debug] [Thread-1 ]: Using redshift connection
Could someone who has successfully implemented it in Redshift share its specific configuration?