Can't specify snapshots on Google Compute

[spuder]

Currently, it appears that you can only create instances on google compute using images.
Terraform should also support using snapshots.

resource "google_compute_instance" "kafka01" {
    name = "kafka01"
    machine_type = "g1-small"
    zone = "us-central1-a"

    disk {
#        image = "ubuntu-1404-trusty-v20141212"
        image = "chef-installed"
    }
    network {
        source = "default"
    }
}

$ terraform apply
google_compute_instance.kafka01: Creating...
  can_ip_forward:             "" => "0"
  disk.#:                     "" => "1"
  disk.0.image:               "" => "chef-installed"
  machine_type:               "" => "g1-small"
  metadata_fingerprint:       "" => "<computed>"
  name:                       "" => "kafka01"
  network.#:                  "" => "1"
  network.0.external_address: "" => "<computed>"
  network.0.internal_address: "" => "<computed>"
  network.0.name:             "" => "<computed>"
  network.0.source:           "" => "default"
  tags_fingerprint:           "" => "<computed>"
  zone:                       "" => "us-central1-a"
google_compute_instance.kafka01: Error: Error loading image 'chef-installed': googleapi: Error 404: The resource 'projects/foobar/global/images/chef-installed' was not found, notFound
Error applying plan:

1 error(s) occurred:

* Error loading image 'chef-installed': googleapi: Error 404: The resource 'projects/brave-reason-831/global/images/chef-installed' was not found, notFound

Terraform does not automatically rollback in the face of errors.
Instead, your Terraform state file has been partially updated with
any resources that successfully completed. Please address the error
above and apply again to incrementally change your infrastructure.

[sparkprime]

The underlying instance API does not allow creating a disk from a snapshot, only an image. In fact it only allows creating the first disk (the boot disk) from an image. The rest have to be existing disks. The Cloud Console actually implemented the operation you have screenshotted as two POST requests, which you can see by clicking on the "Equivalent REST" link at the bottom. That is not the case if you initialize the disk from an image as there is a special sourceImage field in the create instance API.

We could replicate the console behavior, but I'd actually rather not as I see Terraform mirroring the underlying resources in a 1:1 fashion (the Google commandline utility gcloud also does not implement this convenience feature).

I will submit a PR to allow something like the following, would that be sufficient for you?

resource "google_compute_disk" "kafka01" {
    name = "kafka01"
    snapshot = "chef-installed"
}

resource "google_compute_instance" "kafka01" {
    name = "kafka01"
    machine_type = "g1-small"
    zone = "us-central1-a"

    disk {
        disk = "kafka01"
        auto_delete = false
    }
    network {
        source = "default"
    }
}

[spuder]

That would be great, thank you.

[mitchellh]

Fixed by #1426

[kamaljeetrathi]

how to give dynamic name to the instance id

[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 have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.