This message was deleted.
# getting-started
s
This message was deleted.
s
Hi Viktor - I believe the input you want to explicitly set is here: https://www.pulumi.com/registry/packages/gcp/api-docs/sql/databaseinstance/#deletionprotection_nodejs and there's a note in the top description that covers your experience, I believe: https://www.pulumi.com/registry/packages/gcp/api-docs/sql/databaseinstance/#databaseinstance
r
Hi @shy-arm-32391 I use Bigquery, not Cloud SQL. Your link about is about Cloud SQL. But maybe there is a same thing, I'm going to check. I use this resource: https://www.pulumi.com/registry/packages/gcp/api-docs/bigquery/
s
ah, my bad! it does seem that it is the same: https://www.pulumi.com/registry/packages/gcp/api-docs/bigquery/table/
I hope that helps! 🙂
r
It works! When combined with
delete_before_replace=True
. Thanks so much!
b
Hi guys, sorry to chime in that topic but I couldn’t get answer from other channels. @rough-oyster-77458, Do you know if there is a way to add columns a table in Bigquery without recreate table and not touching the underlying data?
r
Hi @breezy-glass-7721 No problem. I use
pulumi_gcp.bigquery.Table
object to create and update tables. It consumes
schema
parameter to control the schema, where we should put a JSON string with a schema. When I just add a new column schema to the JSON, it creates a new column in the existing table without recreating the table. And it works pretty fine for me. But if you want to change a column (for example, when you decide you want to change
user_id
field type from STRING to INT), I'm afraid Pulumi will recreate the whole table. And this is a limitation of Bigquery, as well as most other column oriented data bases. It's not a limitation of Pulumi.
b
Thank you so much Viktor! It is really helpful
We use native api for GCP. Does it make difference?
r
Not much, I guess
But be careful - the native provider is not stable