Feature: Allow users to pass their own IAM roles (#44)

* Feature: Allow users to pass their own IAM roles

* Rename variables following best practices

* update wording for variables & remove unused variable

Author: tirumerla

.github/CODEOWNERS

@@ -15,8 +15,8 @@
 
 # Cloud Posse must review any changes to standard context definition,
 # but some changes can be rubber-stamped.
-**/*.tf       @cloudposse/engineering @cloudposse/approvers
-README.yaml   @cloudposse/engineering @cloudposse/approvers
+**/*.tf       @cloudposse/engineering @cloudposse/contributors @cloudposse/approvers
+README.yaml   @cloudposse/engineering @cloudposse/contributors @cloudposse/approvers
 README.md     @cloudposse/engineering @cloudposse/contributors @cloudposse/approvers
 docs/*.md     @cloudposse/engineering @cloudposse/contributors @cloudposse/approvers
 

.github/auto-release.yml

@@ -17,6 +17,7 @@ version-resolver:
     - 'bugfix'
     - 'bug'
     - 'hotfix'
+    - 'no-release'
   default: 'minor'
 
 categories:
@@ -46,7 +47,7 @@ template: |
 
 replacers:
 # Remove irrelevant information from Renovate bot
-- search: '/---\s+^#.*Renovate configuration(?:.|\n)*?This PR has been generated .*/gm'
+- search: '/(?<=---\s+)+^#.*(Renovate configuration|Configuration)(?:.|\n)*?This PR has been generated .*/gm'
   replace: ''
 # Remove Renovate bot banner image
 - search: '/\[!\[[^\]]*Renovate\][^\]]*\](\([^)]*\))?\s*\n+/gm'

.github/mergify.yml

@@ -56,3 +56,10 @@ pull_request_rules:
       changes_requested: true
       approved: true
       message: "This Pull Request has been updated, so we're dismissing all reviews."
+
+- name: "close Pull Requests without files changed"
+  conditions:
+    - "#files=0"
+  actions:
+    close:
+      message: "This pull request has been automatically closed by Mergify because there are no longer any changes."

.github/workflows/auto-format.yml

@@ -6,7 +6,7 @@ on:
 jobs:
   auto-format:
     runs-on: ubuntu-latest
-    container: cloudposse/build-harness:slim-latest
+    container: cloudposse/build-harness:latest
     steps:
     # Checkout the pull request branch
     #  "An action in a workflow run can’t trigger a new workflow run. For example, if an action pushes code using
@@ -29,6 +29,8 @@ jobs:
     - name: Auto Format
       if: github.event.pull_request.state == 'open'
       shell: bash
+      env:
+        GITHUB_TOKEN: "${{ secrets.PUBLIC_REPO_ACCESS_TOKEN }}"
       run: make BUILD_HARNESS_PATH=/build-harness PACKAGES_PREFER_HOST=true -f /build-harness/templates/Makefile.build-harness pr/auto-format/host
 
     # Commit changes (if any) to the PR branch

.github/workflows/auto-release.yml

@@ -3,17 +3,24 @@ name: auto-release
 on:
   push:
     branches:
-    - master
+      - main
+      - master
+      - production
 
 jobs:
   publish:
     runs-on: ubuntu-latest
     steps:
-    # Drafts your next Release notes as Pull Requests are merged into "master"
-    - uses: release-drafter/release-drafter@v5
-      with:
-        publish: true
-        prerelease: false
-        config-name: auto-release.yml
-      env:
-        GITHUB_TOKEN: ${{ secrets.PUBLIC_REPO_ACCESS_TOKEN }}
+      # Get PR from merged commit to master
+      - uses: actions-ecosystem/action-get-merged-pull-request@v1
+        id: get-merged-pull-request
+        with:
+          github_token: ${{ secrets.PUBLIC_REPO_ACCESS_TOKEN }}
+      # Drafts your next Release notes as Pull Requests are merged into "main"
+      - uses: release-drafter/release-drafter@v5
+        with:
+          publish: ${{ !contains(steps.get-merged-pull-request.outputs.labels, 'no-release') }}
+          prerelease: false
+          config-name: auto-release.yml
+        env:
+          GITHUB_TOKEN: ${{ secrets.PUBLIC_REPO_ACCESS_TOKEN }}

.github/workflows/validate-codeowners.yml

@@ -1,5 +1,7 @@
 name: Validate Codeowners
 on:
+  workflow_dispatch:
+
   pull_request:
 
 jobs:

README.md

@@ -297,8 +297,13 @@ Available targets:
 | <a name="input_delimiter"></a> [delimiter](#input\_delimiter) | Delimiter to be used between ID elements.<br>Defaults to `-` (hyphen). Set to `""` to use no delimiter at all. | `string` | `null` | no |
 | <a name="input_descriptor_formats"></a> [descriptor\_formats](#input\_descriptor\_formats) | Describe additional descriptors to be output in the `descriptors` output map.<br>Map of maps. Keys are names of descriptors. Values are maps of the form<br>`{<br>   format = string<br>   labels = list(string)<br>}`<br>(Type is `any` so the map values can later be enhanced to provide additional options.)<br>`format` is a Terraform format string to be passed to the `format()` function.<br>`labels` is a list of labels, in order, to pass to `format()` function.<br>Label values will be normalized before being passed to `format()` so they will be<br>identical to how they appear in `id`.<br>Default is `{}` (`descriptors` output will be empty). | `any` | `{}` | no |
 | <a name="input_ebs_root_volume_size"></a> [ebs\_root\_volume\_size](#input\_ebs\_root\_volume\_size) | Size in GiB of the EBS root device volume of the Linux AMI that is used for each EC2 instance. Available in Amazon EMR version 4.x and later | `number` | `10` | no |
+| <a name="input_ec2_autoscaling_role_enabled"></a> [ec2\_autoscaling\_role\_enabled](#input\_ec2\_autoscaling\_role\_enabled) | If set to `false`, will use `existing_ec2_autoscaling_role_arn` for an existing EC2 autoscaling IAM role that was created outside of this module | `bool` | `true` | no |
+| <a name="input_ec2_role_enabled"></a> [ec2\_role\_enabled](#input\_ec2\_role\_enabled) | If set to `false`, will use `existing_ec2_instance_profile_arn` for an existing EC2 IAM role that was created outside of this module | `bool` | `true` | no |
 | <a name="input_enabled"></a> [enabled](#input\_enabled) | Set to false to prevent the module from creating any resources | `bool` | `null` | no |
 | <a name="input_environment"></a> [environment](#input\_environment) | ID element. Usually used for region e.g. 'uw2', 'us-west-2', OR role 'prod', 'staging', 'dev', 'UAT' | `string` | `null` | no |
+| <a name="input_existing_ec2_autoscaling_role_arn"></a> [existing\_ec2\_autoscaling\_role\_arn](#input\_existing\_ec2\_autoscaling\_role\_arn) | ARN of an existing EC2 autoscaling role to attach to the cluster | `string` | `""` | no |
+| <a name="input_existing_ec2_instance_profile_arn"></a> [existing\_ec2\_instance\_profile\_arn](#input\_existing\_ec2\_instance\_profile\_arn) | ARN of an existing EC2 instance profile | `string` | `""` | no |
+| <a name="input_existing_service_role_arn"></a> [existing\_service\_role\_arn](#input\_existing\_service\_role\_arn) | ARN of an existing EMR service role to attach to the cluster | `string` | `""` | no |
 | <a name="input_id_length_limit"></a> [id\_length\_limit](#input\_id\_length\_limit) | Limit `id` to this many characters (minimum 6).<br>Set to `0` for unlimited length.<br>Set to `null` for keep the existing setting, which defaults to `0`.<br>Does not affect `id_full`. | `number` | `null` | no |
 | <a name="input_keep_job_flow_alive_when_no_steps"></a> [keep\_job\_flow\_alive\_when\_no\_steps](#input\_keep\_job\_flow\_alive\_when\_no\_steps) | Switch on/off run cluster with no steps or when all steps are complete | `bool` | `true` | no |
 | <a name="input_kerberos_ad_domain_join_password"></a> [kerberos\_ad\_domain\_join\_password](#input\_kerberos\_ad\_domain\_join\_password) | The Active Directory password for ad\_domain\_join\_user. Terraform cannot perform drift detection of this configuration. | `string` | `null` | no |
@@ -334,6 +339,7 @@ Available targets:
 | <a name="input_scale_down_behavior"></a> [scale\_down\_behavior](#input\_scale\_down\_behavior) | The way that individual Amazon EC2 instances terminate when an automatic scale-in activity occurs or an instance group is resized | `string` | `null` | no |
 | <a name="input_security_configuration"></a> [security\_configuration](#input\_security\_configuration) | The security configuration name to attach to the EMR cluster. Only valid for EMR clusters with `release_label` 4.8.0 or greater. See https://www.terraform.io/docs/providers/aws/r/emr_security_configuration.html for more info | `string` | `null` | no |
 | <a name="input_service_access_security_group"></a> [service\_access\_security\_group](#input\_service\_access\_security\_group) | The name of the existing additional security group that will be used for EMR core & task nodes. If empty, a new security group will be created | `string` | `""` | no |
+| <a name="input_service_role_enabled"></a> [service\_role\_enabled](#input\_service\_role\_enabled) | If set to `false`, will use `existing_service_role_arn` for an existing IAM role that was created outside of this module | `bool` | `true` | no |
 | <a name="input_slave_allowed_cidr_blocks"></a> [slave\_allowed\_cidr\_blocks](#input\_slave\_allowed\_cidr\_blocks) | List of CIDR blocks to be allowed to access the slave instances | `list(string)` | `[]` | no |
 | <a name="input_slave_allowed_security_groups"></a> [slave\_allowed\_security\_groups](#input\_slave\_allowed\_security\_groups) | List of security groups to be allowed to connect to the slave instances | `list(string)` | `[]` | no |
 | <a name="input_stage"></a> [stage](#input\_stage) | ID element. Usually used to indicate role, e.g. 'prod', 'staging', 'source', 'build', 'test', 'deploy', 'release' | `string` | `null` | no |

docs/terraform.md

@@ -91,8 +91,13 @@
 | <a name="input_delimiter"></a> [delimiter](#input\_delimiter) | Delimiter to be used between ID elements.<br>Defaults to `-` (hyphen). Set to `""` to use no delimiter at all. | `string` | `null` | no |
 | <a name="input_descriptor_formats"></a> [descriptor\_formats](#input\_descriptor\_formats) | Describe additional descriptors to be output in the `descriptors` output map.<br>Map of maps. Keys are names of descriptors. Values are maps of the form<br>`{<br>   format = string<br>   labels = list(string)<br>}`<br>(Type is `any` so the map values can later be enhanced to provide additional options.)<br>`format` is a Terraform format string to be passed to the `format()` function.<br>`labels` is a list of labels, in order, to pass to `format()` function.<br>Label values will be normalized before being passed to `format()` so they will be<br>identical to how they appear in `id`.<br>Default is `{}` (`descriptors` output will be empty). | `any` | `{}` | no |
 | <a name="input_ebs_root_volume_size"></a> [ebs\_root\_volume\_size](#input\_ebs\_root\_volume\_size) | Size in GiB of the EBS root device volume of the Linux AMI that is used for each EC2 instance. Available in Amazon EMR version 4.x and later | `number` | `10` | no |
+| <a name="input_ec2_autoscaling_role_enabled"></a> [ec2\_autoscaling\_role\_enabled](#input\_ec2\_autoscaling\_role\_enabled) | If set to `false`, will use `existing_ec2_autoscaling_role_arn` for an existing EC2 autoscaling IAM role that was created outside of this module | `bool` | `true` | no |
+| <a name="input_ec2_role_enabled"></a> [ec2\_role\_enabled](#input\_ec2\_role\_enabled) | If set to `false`, will use `existing_ec2_instance_profile_arn` for an existing EC2 IAM role that was created outside of this module | `bool` | `true` | no |
 | <a name="input_enabled"></a> [enabled](#input\_enabled) | Set to false to prevent the module from creating any resources | `bool` | `null` | no |
 | <a name="input_environment"></a> [environment](#input\_environment) | ID element. Usually used for region e.g. 'uw2', 'us-west-2', OR role 'prod', 'staging', 'dev', 'UAT' | `string` | `null` | no |
+| <a name="input_existing_ec2_autoscaling_role_arn"></a> [existing\_ec2\_autoscaling\_role\_arn](#input\_existing\_ec2\_autoscaling\_role\_arn) | ARN of an existing EC2 autoscaling role to attach to the cluster | `string` | `""` | no |
+| <a name="input_existing_ec2_instance_profile_arn"></a> [existing\_ec2\_instance\_profile\_arn](#input\_existing\_ec2\_instance\_profile\_arn) | ARN of an existing EC2 instance profile | `string` | `""` | no |
+| <a name="input_existing_service_role_arn"></a> [existing\_service\_role\_arn](#input\_existing\_service\_role\_arn) | ARN of an existing EMR service role to attach to the cluster | `string` | `""` | no |
 | <a name="input_id_length_limit"></a> [id\_length\_limit](#input\_id\_length\_limit) | Limit `id` to this many characters (minimum 6).<br>Set to `0` for unlimited length.<br>Set to `null` for keep the existing setting, which defaults to `0`.<br>Does not affect `id_full`. | `number` | `null` | no |
 | <a name="input_keep_job_flow_alive_when_no_steps"></a> [keep\_job\_flow\_alive\_when\_no\_steps](#input\_keep\_job\_flow\_alive\_when\_no\_steps) | Switch on/off run cluster with no steps or when all steps are complete | `bool` | `true` | no |
 | <a name="input_kerberos_ad_domain_join_password"></a> [kerberos\_ad\_domain\_join\_password](#input\_kerberos\_ad\_domain\_join\_password) | The Active Directory password for ad\_domain\_join\_user. Terraform cannot perform drift detection of this configuration. | `string` | `null` | no |
@@ -128,6 +133,7 @@
 | <a name="input_scale_down_behavior"></a> [scale\_down\_behavior](#input\_scale\_down\_behavior) | The way that individual Amazon EC2 instances terminate when an automatic scale-in activity occurs or an instance group is resized | `string` | `null` | no |
 | <a name="input_security_configuration"></a> [security\_configuration](#input\_security\_configuration) | The security configuration name to attach to the EMR cluster. Only valid for EMR clusters with `release_label` 4.8.0 or greater. See https://www.terraform.io/docs/providers/aws/r/emr_security_configuration.html for more info | `string` | `null` | no |
 | <a name="input_service_access_security_group"></a> [service\_access\_security\_group](#input\_service\_access\_security\_group) | The name of the existing additional security group that will be used for EMR core & task nodes. If empty, a new security group will be created | `string` | `""` | no |
+| <a name="input_service_role_enabled"></a> [service\_role\_enabled](#input\_service\_role\_enabled) | If set to `false`, will use `existing_service_role_arn` for an existing IAM role that was created outside of this module | `bool` | `true` | no |
 | <a name="input_slave_allowed_cidr_blocks"></a> [slave\_allowed\_cidr\_blocks](#input\_slave\_allowed\_cidr\_blocks) | List of CIDR blocks to be allowed to access the slave instances | `list(string)` | `[]` | no |
 | <a name="input_slave_allowed_security_groups"></a> [slave\_allowed\_security\_groups](#input\_slave\_allowed\_security\_groups) | List of security groups to be allowed to connect to the slave instances | `list(string)` | `[]` | no |
 | <a name="input_stage"></a> [stage](#input\_stage) | ID element. Usually used to indicate role, e.g. 'prod', 'staging', 'source', 'build', 'test', 'deploy', 'release' | `string` | `null` | no |

main.tf

@@ -264,7 +264,7 @@ This role is required for all clusters.
 https://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-iam-roles.html
 */
 data "aws_iam_policy_document" "assume_role_emr" {
-  count = module.this.enabled ? 1 : 0
+  count = module.this.enabled && var.service_role_enabled ? 1 : 0
 
   statement {
     effect = "Allow"
@@ -279,7 +279,7 @@ data "aws_iam_policy_document" "assume_role_emr" {
 }
 
 resource "aws_iam_role" "emr" {
-  count              = module.this.enabled ? 1 : 0
+  count              = module.this.enabled && var.service_role_enabled ? 1 : 0
   name               = module.label_emr.id
   assume_role_policy = join("", data.aws_iam_policy_document.assume_role_emr.*.json)
 
@@ -288,7 +288,7 @@ resource "aws_iam_role" "emr" {
 
 # https://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-iam-roles.html
 resource "aws_iam_role_policy_attachment" "emr" {
-  count      = module.this.enabled ? 1 : 0
+  count      = module.this.enabled && var.service_role_enabled ? 1 : 0
   role       = join("", aws_iam_role.emr.*.name)
   policy_arn = "arn:aws:iam::aws:policy/service-role/AmazonElasticMapReduceRole"
 }
@@ -301,7 +301,7 @@ This role is required for all clusters.
 https://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-iam-roles.html
 */
 data "aws_iam_policy_document" "assume_role_ec2" {
-  count = module.this.enabled ? 1 : 0
+  count = module.this.enabled && var.ec2_role_enabled ? 1 : 0
 
   statement {
     effect = "Allow"
@@ -316,7 +316,7 @@ data "aws_iam_policy_document" "assume_role_ec2" {
 }
 
 resource "aws_iam_role" "ec2" {
-  count              = module.this.enabled ? 1 : 0
+  count              = module.this.enabled && var.ec2_role_enabled ? 1 : 0
   name               = module.label_ec2.id
   assume_role_policy = join("", data.aws_iam_policy_document.assume_role_ec2.*.json)
 
@@ -325,13 +325,13 @@ resource "aws_iam_role" "ec2" {
 
 # https://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-iam-roles.html
 resource "aws_iam_role_policy_attachment" "ec2" {
-  count      = module.this.enabled ? 1 : 0
+  count      = module.this.enabled && var.ec2_role_enabled ? 1 : 0
   role       = join("", aws_iam_role.ec2.*.name)
   policy_arn = "arn:aws:iam::aws:policy/service-role/AmazonElasticMapReduceforEC2Role"
 }
 
 resource "aws_iam_instance_profile" "ec2" {
-  count = module.this.enabled ? 1 : 0
+  count = module.this.enabled && var.ec2_role_enabled ? 1 : 0
   name  = join("", aws_iam_role.ec2.*.name)
   role  = join("", aws_iam_role.ec2.*.name)
 }
@@ -342,7 +342,7 @@ This role is required for all clusters.
 https://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-iam-roles.html
 */
 resource "aws_iam_role" "ec2_autoscaling" {
-  count              = module.this.enabled ? 1 : 0
+  count              = module.this.enabled && var.ec2_autoscaling_role_enabled ? 1 : 0
   name               = module.label_ec2_autoscaling.id
   assume_role_policy = join("", data.aws_iam_policy_document.assume_role_emr.*.json)
 
@@ -351,7 +351,7 @@ resource "aws_iam_role" "ec2_autoscaling" {
 
 # https://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-iam-roles.html
 resource "aws_iam_role_policy_attachment" "ec2_autoscaling" {
-  count      = module.this.enabled ? 1 : 0
+  count      = module.this.enabled && var.ec2_autoscaling_role_enabled ? 1 : 0
   role       = join("", aws_iam_role.ec2_autoscaling.*.name)
   policy_arn = "arn:aws:iam::aws:policy/service-role/AmazonElasticMapReduceforAutoScalingRole"
 }
@@ -392,7 +392,7 @@ resource "aws_emr_cluster" "default" {
     emr_managed_master_security_group = var.use_existing_managed_master_security_group == false ? join("", aws_security_group.managed_master.*.id) : var.managed_master_security_group
     emr_managed_slave_security_group  = var.use_existing_managed_slave_security_group == false ? join("", aws_security_group.managed_slave.*.id) : var.managed_slave_security_group
     service_access_security_group     = var.use_existing_service_access_security_group == false && var.subnet_type == "private" ? join("", aws_security_group.managed_service_access.*.id) : var.service_access_security_group
-    instance_profile                  = join("", aws_iam_instance_profile.ec2.*.arn)
+    instance_profile                  = var.ec2_role_enabled ? join("", aws_iam_instance_profile.ec2.*.arn) : var.existing_ec2_instance_profile_arn
     additional_master_security_groups = var.use_existing_additional_master_security_group == false ? join("", aws_security_group.master.*.id) : var.additional_master_security_group
     additional_slave_security_groups  = var.use_existing_additional_slave_security_group == false ? join("", aws_security_group.slave.*.id) : var.additional_slave_security_group
   }
@@ -479,8 +479,8 @@ resource "aws_emr_cluster" "default" {
 
   log_uri = var.log_uri
 
-  service_role     = join("", aws_iam_role.emr.*.arn)
-  autoscaling_role = join("", aws_iam_role.ec2_autoscaling.*.arn)
+  service_role     = var.service_role_enabled ? join("", aws_iam_role.emr.*.arn) : var.existing_service_role_arn
+  autoscaling_role = var.ec2_autoscaling_role_enabled ? join("", aws_iam_role.ec2_autoscaling.*.arn) : var.existing_ec2_autoscaling_role_arn
 
   tags = module.this.tags
 

outputs.tf

@@ -29,6 +29,6 @@ output "master_host" {
 }
 
 output "ec2_role" {
-  value       = join("", aws_iam_role.ec2.*.name)
+  value       = var.ec2_role_enabled ? join("", aws_iam_role.ec2.*.name) : null
   description = "Role name of EMR EC2 instances so users can attach more policies"
 }