provider/digitalocean: prevent new resources when using ID's of image…

[roidelapluie]

…s with slugs

When you specify the ID of an image that has a slug, terraform would
store its slug to the state, hence it would always recreate the image.

This commit fixes it by storing the image as an ID when it is specified
by and ID by the user, ignoring the slug.

Closes #12751.
Fixes #12255.

Signed-off-by: Julien Pivotto roidelapluie@inuits.eu

[roidelapluie]

ACCEPTANCE TESTS WITH TESTS ONLY (test_droplet.go)

==> Checking that code complies with gofmt requirements...
go generate $(go list ./... | grep -v /terraform/vendor/)
2017/04/23 18:04:19 Generated command/internal_plugin_list.go
TF_ACC=1 go test ./builtin/providers/digitalocean -v -run=TestAccDigitalOceanDroplet_WithID -timeout 120m
=== RUN   TestAccDigitalOceanDroplet_WithID
--- FAIL: TestAccDigitalOceanDroplet_WithID (38.42s)
        testing.go:280: Step 0 error: After applying this step, the plan was not empty:

                DIFF:

                DESTROY/CREATE: digitalocean_droplet.foobar
                  disk:                 "20" => "<computed>"
                  image:                "centos-7-x64" => "22995941" (forces new resource)
                  ipv4_address:         "45.55.154.238" => "<computed>"
                  ipv4_address_private: "" => "<computed>"
                  ipv6_address:         "" => "<computed>"
                  ipv6_address_private: "" => "<computed>"
                  locked:               "false" => "<computed>"
                  name:                 "foo-4889334322092258525" => "foo-4889334322092258525"
                  price_hourly:         "0.00744" => "<computed>"
                  price_monthly:        "5" => "<computed>"
                  region:               "nyc3" => "nyc3"
                  resize_disk:          "true" => "true"
                  size:                 "512mb" => "512mb"
                  status:               "active" => "<computed>"
                  user_data:            "foobar" => "foobar"
                  vcpus:                "1" => "<computed>"

                STATE:

                digitalocean_droplet.foobar:
                  ID = 46676585
                  disk = 20
                  image = centos-7-x64
                  ipv4_address = 45.55.154.238
                  locked = false
                  name = foo-4889334322092258525
                  price_hourly = 0.00744
                  price_monthly = 5
                  region = nyc3
                  resize_disk = true
                  size = 512mb
                  status = active
                  tags.# = 0
                  user_data = foobar
                  vcpus = 1
FAIL
exit status 1
FAIL    github.com/hashicorp/terraform/builtin/providers/digitalocean   38.429s
make: *** [Makefile:49: testacc] Error 1

ACCEPTANCE TESTS WITH ALL CHANGES (droplet_test.go + droplet.go)

make testacc TEST=./builtin/providers/digitalocean TESTARGS='-run=TestAccDigitalOceanDroplet_WithID'  
==> Checking that code complies with gofmt requirements...
go generate $(go list ./... | grep -v /terraform/vendor/)
2017/04/23 18:02:16 Generated command/internal_plugin_list.go
TF_ACC=1 go test ./builtin/providers/digitalocean -v -run=TestAccDigitalOceanDroplet_WithID -timeout 120m
=== RUN   TestAccDigitalOceanDroplet_WithID
--- PASS: TestAccDigitalOceanDroplet_WithID (50.56s)
PASS
ok      github.com/hashicorp/terraform/builtin/providers/digitalocean   50.564s

I also ran the basic test:

make testacc TEST=./builtin/providers/digitalocean TESTARGS='-run=TestAccDigitalOceanDroplet_Basic'
==> Checking that code complies with gofmt requirements...
go generate $(go list ./... | grep -v /terraform/vendor/)
2017/04/23 18:16:05 Generated command/internal_plugin_list.go
TF_ACC=1 go test ./builtin/providers/digitalocean -v -run=TestAccDigitalOceanDroplet_Basic -timeout 120m
=== RUN   TestAccDigitalOceanDroplet_Basic
--- PASS: TestAccDigitalOceanDroplet_Basic (49.30s)
PASS
ok      github.com/hashicorp/terraform/builtin/providers/digitalocean   49.308s

[roidelapluie]

There might be a problem when switching from a snapshot to a distro

[roidelapluie]

fixed.

[roidelapluie]

current status: ready for review & merge

[radeksimko]

Just one minor thing in the test, otherwise this LGTM.

[radeksimko]

We actually refresh the state as part of every test step, so this is not necessary - would you mind removing the second step + Destroy: false?

[roidelapluie]

done

[radeksimko]

@roidelapluie thanks for sending the other PR, would you mind modifying this PR/test too and resolving conflicts?

Thanks

[roidelapluie]

done

[radeksimko]

👍 Great, thanks for the short response time, pulling in...

[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.