Creating HA google_redis_instance doesn't seem to store auto-generated alternative_location_id

[reidab]

Terraform Version

Terraform v0.11.7
+ provider.google v1.14.0

Affected Resource(s)

• google_redis_instance

Terraform Configuration Files

resource "google_redis_instance" "production" {
  name = "redis-production"
  memory_size_gb = 1
  tier = "STANDARD_HA"
  region = "us-central1"
  location_id = "us-central1-a"
  authorized_network = "${google_compute_network.the_dyrt.self_link}"
}

Expected Behavior

When creating a google_redis_instance without specifying alternative_location_id, the alternative location is automatically assigned. I'd expect this computed value to be persisted to the state in such a way that subsequent updates didn't try to clear it.

Actual Behavior

On subsequent terraform apply runs with no other related changes, terraform wants to recreate the instance to set alternative_location_id back to a blank string:

Terraform will perform the following actions:

-/+ module.google_cloud.google_redis_instance.production (new resource required)
      id:                          "the-dyrt/us-central1/redis-production" => <computed> (forces new resource)
      alternative_location_id:     "us-central1-b" => "" (forces new resource)
      authorized_network:          "projects/the-dyrt/global/networks/the-dyrt" => "https://www.googleapis.com/compute/v1/projects/the-dyrt/global/networks/the-dyrt"
      create_time:                 "2018-06-18T22:26:15.486557Z" => <computed>
      current_location_id:         "us-central1-a" => <computed>
      host:                        "10.18.0.12" => <computed>
      location_id:                 "us-central1-a" => "us-central1-a"
      memory_size_gb:              "1" => "1"
      name:                        "redis-production" => "redis-production"
      port:                        "6379" => <computed>
      project:                     "the-dyrt" => <computed>
      redis_version:               "REDIS_3_2" => <computed>
      region:                      "us-central1" => "us-central1"
      reserved_ip_range:           "10.18.0.8/29" => <computed>
      tier:                        "STANDARD_HA" => "STANDARD_HA"

Manually modifying the instance specification and setting alternative_location_id = "us-central1-b" to match the generated value works around this and terraform no longer desires to recreate the instance.

Steps to Reproduce

1. Specify a STANDARD_HA redis instance without an explicit alternative_location_id
2. terraform apply
3. terraform apply

[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!