provider/aws: export elasticache nodes

[phinze]

Here's a start to a fix for #1918

Needs to wait for len(cluster.CacheNodes) == cluster.NumCacheNodes, since
apparently that takes a bit of time and the initial response always has
an empty collection of nodes.

I think that doing a complex Computed attribute is okay, and you can reference it like aws_elasticache_cluster.foo.cache_nodes.0.address.

pushing up the WIP in case @catsby has bandwidth to finish this first

---

@catsby is doing:

• export cache node info (aws_elasticache_cluster does not export node endpoints or IP addresses #1918)
• touch ups (don't require parameter_group, others)
• add tag support (aws_elasticache_cluster does not support the tags attribute #1919) (d8f3783)

[phinze]

Ooo nice catch.

[catsby]

Needs to wait for len(cluster.CacheNodes) == cluster.NumCacheNodes, since
apparently that takes a bit of time and the initial response always has
an empty collection of nodes.

You need to explicitly tell the API that you want node info with ShowCacheNodeInfo: true

[phinze]

Cool cool - LGTM!

[saulshanabrook]

@catsby Thanks so much!

I am currently having a problem where sometimes terraform doesn't wait for the cluster to come up completely before continuing, so that its nodes don't exist.

The state looks like this:

        {
            "path": [
                "root",
                "redis-reports"
            ],
            "outputs": {
                "port": "0"
            },
            "resources": {
                "aws_elasticache_cluster.default": {
                    "type": "aws_elasticache_cluster",
                    "primary": {
                        "id": "redis-reports",
                        "attributes": {
                            "cache_nodes.#": "0",
                            "cluster_id": "redis-reports",
                            "engine": "redis",
                            "engine_version": "2.8.19",
                            "id": "redis-reports",
                            "node_type": "cache.t2.micro",
                            "num_cache_nodes": "1",
                            "port": "11211",
                            "security_group_ids.#": "0",
                            "security_group_names.#": "0",
                            "subnet_group_name": "default-new",
                            "tags.#": "0"
                        }
                    }
                }
            }
        }

So terraform apply will fail sometimes because it tries to then get the first node and fails. If I wait until the node comes up on amazon and then run terraform refresh, the state is populated fine.

[catsby]

So terraform apply will fail sometimes because it tries to then get the first node and fails.

I'm not sure I follow.. do you have any error messages you can share?

[saulshanabrook]

@catsby

Error applying plan:

1 error(s) occurred:

* 1 error(s) occurred:

* Resource 'aws_elasticache_cluster.default' does not have attribute 'cache_nodes.0.address' for variable 'aws_elasticache_cluster.default.cache_nodes.0.address'

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.

That's because I reference 'aws_elasticache_cluster.default.cache_nodes.0.address afterword and the variable doesn't exist because elasticache isn't really done provisioning the cluster.

[saulshanabrook]

Then I wait for the cluster to finish provisioning, run terraform refresh and then terraform apply, and it will run without failing.

[catsby]

Understood, thanks! Phinze above suspected this, I'll check it out in the
morning

On Thursday, May 14, 2015, Saul Shanabrook notifications@github.com wrote:

Then I wait for the cluster to finish provisioning, run terraform refresh
and then terraform apply, and it will run without failing.

—
Reply to this email directly or view it on GitHub
#1965 (comment).

Clint

[juniorplenty]

@catsby I'm seeing the same race condition and error in 0.5.2, any update on this?

[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.