google_container_cluster fields recreate the cluster unnecesarily

[paddycarver]

It should be possible to update, e.g., the node_pool field on google_container_cluster without tearing down the cluster and recreating it. However, at the moment, Terraform requires the cluster to be destroyed and recreated in order to make changes to many of the fields.

[paddycarver]

When google_container_cluster was created, the API didn't support updating most (or all, I forget) of its fields. Now the API has progressed, but the resource hasn't been updated to support the new functionality. I think for a lot of this, it's just a matter of updating the resource to do the update API call and marking the fields as ForceNew: false.

[davidquarles]

Realistically, this behavior (specifically: when mutating google_container_cluster.node_pool) is the main deal-breaker for us (with IP Aliases as a close second), and I'd really like to use terraform to keep our tooling consistent with cluster provisioning on the AWS side. If you'd like help, I have bandwidth to do so this week.

[danawillow]

@davidquarles happy to have help! Just let me know which ones you're planning on doing so I don't accidentally do the same ones.

[danawillow]

Closing this in favor of more granular issues per field that needs to be updatable (I think most are now anyway)

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. If you feel I made an error 🤖 🙉 , please reach out to my human friends 👉 hashibot-feedback@hashicorp.com. Thanks!