Can't interpolate multi resources for nested fields

[hashibot]

This issue was originally opened by @calvinfo as hashicorp/terraform#3902. It was migrated here as part of the provider split. The original body of the issue is below.

---

We're attempting to create an elasticache cluster, but unfortunately the interpolation doesn't work for splats of nested fields.

resource "aws_elasticache_cluster" "cache" {
  cluster_id = "api-dedupe-memcached"
  engine = "memcached"
  node_type = "cache.r3.8xlarge"
  port = 11211
  num_cache_nodes = 2
  parameter_group_name = "default.memcached1.4"
  subnet_group_name = "${aws_elasticache_subnet_group.default.name}"
  security_group_ids = ["${aws_security_group.memcached.id}"]
}

# and then elsewhere...

value = "${join(",", aws_elasticache_cluster.cache.cache_nodes.*.address)}"

It could be fixed in the config/interpolation, to mark it as a multi, but that'd require changing around the field name behavior as well. Is that behavior which you're willing to support? Otherwise it might be worth changing the cache resource.

[hashibot]

This comment was originally opened by @erkolson as hashicorp/terraform#3902 (comment). It was migrated here as part of the provider split. The original comment is below.

---

I'm seeing what I assume is the same issue with the element interpolation.

value = ["${elements(aws_elasticache_cluster.cache.cache_nodes.*.address, count.index}"]

[hashibot]

This comment was originally opened by @calvinfo as hashicorp/terraform#3902 (comment). It was migrated here as part of the provider split. The original comment is below.

---

FWIW, the workaround we ended up using was this:

resource "aws_elasticache_cluster" "main" {
  cluster_id = "${var.cluster_id}"
  engine = "${var.engine}"
  node_type = "${var.node_type}"
  port = "${lookup(var.port, var.engine)}"
  num_cache_nodes = "${var.node_count}"
  parameter_group_name = "default.memcached1.4"
  subnet_group_name = "${aws_elasticache_subnet_group.main.name}"
  security_group_ids = ["${aws_security_group.main.id}"]
}

resource "template_file" "hosts" {
  count = "${var.node_count}"
  template = "${file("${path.module}/template.tpl")}"

  vars {
    endpoint = "${aws_elasticache_cluster.main.configuration_endpoint}"
    count = "${format("%04d", count.index + 1)}"
  }
}

/**
 * Outputs.
 */

output "node_endpoints" { value = "${join(",", template_file.hosts.*.rendered)}" }

We created a module for booting up an elasticache, which gets passed a node_count. And then we create multiple template_files that look like this:

${replace("${endpoint}", "cfg", "${count}")}

It's admittedly hacky, since it just relies on how AWS formats their configuration endpoint and then replaces the string to find the addresses of individual nodes. But, it does work, and outputs a list of individual nodes from the module.

[hashibot]

This comment was originally opened by @erkolson as hashicorp/terraform#3902 (comment). It was migrated here as part of the provider split. The original comment is below.

---

Thanks, that is clever. For my needs, this is dirty but will work:

resource "aws_route53_record" "cache_nodes" {
  count = "${var.number_o_cache_nodes}"
  name = "a-predictable-name-cache-node-${format("%04d", count.index + 1)}"
  type = "CNAME"
  records = ["${replace(replace(aws_elasticache_cluster.cache.configuration_endpoint, "cfg", format("%04d", count.index + 1)), ":11211", "")}"]
  ttl = "60"
  zone_id = "${var.my_zone}"
}

[hashibot]

This comment was originally opened by @svenwltr as hashicorp/terraform#3902 (comment). It was migrated here as part of the provider split. The original comment is below.

---

This issue is open for more than a year now. Any updates or plans to fix it?

[hashibot]

This comment was originally opened by @phillbaker as hashicorp/terraform#3902 (comment). It was migrated here as part of the provider split. The original comment is below.

---

Not sure if this is exactly the same underlying issue but it looks like splats for lists of nested fields were working in 0.7.x, for example:

output "url" {
  value = "${aws_elasticache_cluster.redis.cache_nodes.*.address}"
}

There are quite a few references to this on github, 1 and 2 for example.

However, at some point in 0.8.x it looks like this has stopped working and the outputs return nothing now.

[n3ph]

Version

$ terraform --version
Terraform v0.11.7
+ provider.aws v1.25.0
+ provider.null v1.0.0

Explanation

The reason why this is not working as expected is that the HCL interpreter treats splat expressions in some special kind. There are different Issues filed by people having problems with that (#672)

This behavior will change with HCL2...
hashicorp/terraform#17048 (comment)

BTW as of hashicorp/terraform#2821 splats in outputs as lists are supported for a long time now.

Workaround

You may use the null_data_source. As aws_elasticache_cluster.redis.cache_nodes returns a list of maps, we only need to count over the list and map out all desired keys. But to get this work we need to use TF_WARN_OUTPUT_ERRORS=1 .

#------------------------------------------------------------------------------#
# Test Setup
#------------------------------------------------------------------------------#

provider "aws" {
  max_retries = 3
  region      = "eu-central-1"
  profile     = "devops"
}

data "aws_vpc" "default" {
  default = true
}

data "aws_subnet_ids" "default" {
  vpc_id = "${data.aws_vpc.default.id}"
}

data "aws_security_group" "default" {
  vpc_id = "${data.aws_vpc.default.id}"
}

#------------------------------------------------------------------------------#
# Memcache Cluster
#------------------------------------------------------------------------------#

variable "cluster_size" {
  description = "Number of cluster nodes"
  default     = 2
}

resource "aws_elasticache_subnet_group" "test" {
  name       = "test"
  subnet_ids = ["${data.aws_subnet_ids.default.ids}"]
}

resource "aws_elasticache_cluster" "test" {
  cluster_id           = "test"
  engine               = "memcached"
  node_type            = "cache.t2.micro"
  port                 = 11211
  num_cache_nodes      = "${var.cluster_size}"
  parameter_group_name = "default.memcached1.4"
  subnet_group_name    = "${aws_elasticache_subnet_group.test.name}"
  security_group_ids   = ["${data.aws_security_group.default.id}"]
}

#------------------------------------------------------------------------------#
# Output rendering
#------------------------------------------------------------------------------#

data "null_data_source" "test" {
  count = "${var.cluster_size}"

  inputs = {
    address = "${lookup(aws_elasticache_cluster.test.cache_nodes[count.index], "address")}"
  }
}

output "test" {
  value = ["${data.null_data_source.test.*.outputs.address}"]
}

$ TF_WARN_OUTPUT_ERRORS=1 terraform apply
data.aws_vpc.default: Refreshing state...
data.aws_subnet_ids.default: Refreshing state...
data.aws_security_group.default: Refreshing state...

An execution plan has been generated and is shown below.
Resource actions are indicated with the following symbols:
  + create
 <= read (data resources)

Terraform will perform the following actions:

 <= data.null_data_source.test[0]
      id:                            <computed>
      has_computed_default:          <computed>
      inputs.%:                      <computed>
      outputs.%:                     <computed>
      random:                        <computed>

 <= data.null_data_source.test[1]
      id:                            <computed>
      has_computed_default:          <computed>
      inputs.%:                      <computed>
      outputs.%:                     <computed>
      random:                        <computed>

  + aws_elasticache_cluster.test
      id:                            <computed>
      apply_immediately:             <computed>
      availability_zone:             <computed>
      az_mode:                       <computed>
      cache_nodes.#:                 <computed>
      cluster_address:               <computed>
      cluster_id:                    "test"
      configuration_endpoint:        <computed>
      engine:                        "memcached"
      engine_version:                <computed>
      maintenance_window:            <computed>
      node_type:                     "cache.t2.micro"
      num_cache_nodes:               "2"
      parameter_group_name:          "default.memcached1.4"
      port:                          "11211"
      replication_group_id:          <computed>
      security_group_ids.#:          "1"
      security_group_ids.3425856862: "sg-7e92ac14"
      security_group_names.#:        <computed>
      snapshot_window:               <computed>
      subnet_group_name:             "test"

  + aws_elasticache_subnet_group.test
      id:                            <computed>
      description:                   "Managed by Terraform"
      name:                          "test"
      subnet_ids.#:                  "3"
      subnet_ids.1289025745:         "subnet-c128938c"
      subnet_ids.3736608364:         "subnet-daf525a7"
      subnet_ids.4040429980:         "subnet-9fb52cf4"


Plan: 2 to add, 0 to change, 0 to destroy.

Do you want to perform these actions?
  Terraform will perform the actions described above.
  Only 'yes' will be accepted to approve.

  Enter a value: yes

aws_elasticache_subnet_group.test: Creating...
  description:           "" => "Managed by Terraform"
  name:                  "" => "test"
  subnet_ids.#:          "" => "3"
  subnet_ids.1289025745: "" => "subnet-c128938c"
  subnet_ids.3736608364: "" => "subnet-daf525a7"
  subnet_ids.4040429980: "" => "subnet-9fb52cf4"
aws_elasticache_subnet_group.test: Creation complete after 1s (ID: test)
aws_elasticache_cluster.test: Creating...
  apply_immediately:             "" => "<computed>"
  availability_zone:             "" => "<computed>"
  az_mode:                       "" => "<computed>"
  cache_nodes.#:                 "" => "<computed>"
  cluster_address:               "" => "<computed>"
  cluster_id:                    "" => "test"
  configuration_endpoint:        "" => "<computed>"
  engine:                        "" => "memcached"
  engine_version:                "" => "<computed>"
  maintenance_window:            "" => "<computed>"
  node_type:                     "" => "cache.t2.micro"
  num_cache_nodes:               "" => "2"
  parameter_group_name:          "" => "default.memcached1.4"
  port:                          "" => "11211"
  replication_group_id:          "" => "<computed>"
  security_group_ids.#:          "" => "1"
  security_group_ids.3425856862: "" => "sg-7e92ac14"
  security_group_names.#:        "" => "<computed>"
  snapshot_window:               "" => "<computed>"
  subnet_group_name:             "" => "test"
aws_elasticache_cluster.test: Still creating... (10s elapsed)
aws_elasticache_cluster.test: Still creating... (20s elapsed)
aws_elasticache_cluster.test: Still creating... (30s elapsed)
aws_elasticache_cluster.test: Still creating... (40s elapsed)
aws_elasticache_cluster.test: Still creating... (50s elapsed)
aws_elasticache_cluster.test: Still creating... (1m0s elapsed)
aws_elasticache_cluster.test: Still creating... (1m10s elapsed)
aws_elasticache_cluster.test: Still creating... (1m20s elapsed)
aws_elasticache_cluster.test: Still creating... (1m30s elapsed)
aws_elasticache_cluster.test: Still creating... (1m40s elapsed)
aws_elasticache_cluster.test: Still creating... (1m50s elapsed)
aws_elasticache_cluster.test: Still creating... (2m0s elapsed)
aws_elasticache_cluster.test: Still creating... (2m10s elapsed)
aws_elasticache_cluster.test: Still creating... (2m20s elapsed)
aws_elasticache_cluster.test: Still creating... (2m30s elapsed)
aws_elasticache_cluster.test: Still creating... (2m40s elapsed)
aws_elasticache_cluster.test: Still creating... (2m50s elapsed)
aws_elasticache_cluster.test: Still creating... (3m0s elapsed)
aws_elasticache_cluster.test: Still creating... (3m10s elapsed)
aws_elasticache_cluster.test: Still creating... (3m20s elapsed)
aws_elasticache_cluster.test: Still creating... (3m30s elapsed)
aws_elasticache_cluster.test: Still creating... (3m40s elapsed)
aws_elasticache_cluster.test: Still creating... (3m50s elapsed)
aws_elasticache_cluster.test: Still creating... (4m0s elapsed)
aws_elasticache_cluster.test: Still creating... (4m10s elapsed)
aws_elasticache_cluster.test: Still creating... (4m20s elapsed)
aws_elasticache_cluster.test: Still creating... (4m30s elapsed)
aws_elasticache_cluster.test: Creation complete after 4m38s (ID: test)
data.null_data_source.test[1]: Refreshing state...
data.null_data_source.test[0]: Refreshing state...

Apply complete! Resources: 2 added, 0 changed, 0 destroyed.

Outputs:

test = [
    test.b6reqy.0001.euc1.cache.amazonaws.com,
    test.b6reqy.0002.euc1.cache.amazonaws.com
]

[bflad]

Hi folks 👋 This issue is resolved in Terraform 0.12, which fully supports indexed splat (*) notation with list attribute values.

Given this configuration:

terraform {
  required_providers {
    aws = "2.20.0"
  }
  required_version = "0.12.5"
}

provider "aws" {
  region = "us-east-2"
}

resource "aws_elasticache_cluster" "test" {
  cluster_id      = "bflad-testing"
  engine          = "memcached"
  node_type       = "cache.t2.micro"
  num_cache_nodes = 3
}

output "test1" {
  value = aws_elasticache_cluster.test.cache_nodes[*].address
}

output "test2" {
  value = element(aws_elasticache_cluster.test.cache_nodes[*].address, 1)
}

output "test3" {
  value = join(",", formatlist("%s:%s", aws_elasticache_cluster.test.cache_nodes[*].address, aws_elasticache_cluster.test.cache_nodes[*].port))
}

We can get the various expected outputs that were problematic in Terraform 0.11 and earlier:

Outputs:

test1 = [
  "bflad-testing.asxu6y.0001.use2.cache.amazonaws.com",
  "bflad-testing.asxu6y.0002.use2.cache.amazonaws.com",
  "bflad-testing.asxu6y.0003.use2.cache.amazonaws.com",
]
test2 = bflad-testing.asxu6y.0002.use2.cache.amazonaws.com
test3 = bflad-testing.asxu6y.0001.use2.cache.amazonaws.com:11211,bflad-testing.asxu6y.0002.use2.cache.amazonaws.com:11211,bflad-testing.asxu6y.0003.use2.cache.amazonaws.com:11211

Enjoy! 🚀

[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. Thanks!