Add option to specify environment type (#65)

* Add an option to specify environment type

* Update README

* Classic typo

* Update doc

* Update doc some more

* make readme

* More doc updates

* Remove unused settings in SingleInstance environment

They get ignored by ElasticBeanstalk anyway, but they trigger a noisy plan when terraform keeps trying to add them

* Revert these "fixes", they might be wrong

* Document `wait_for_ready_timeout`

* Revert "Remove unused settings in SingleInstance environment"

This reverts commit c9a9cb4

Author: skolsuper

README.md

@@ -90,6 +90,7 @@ Available targets:
 | env_default_key | Default ENV variable key for Elastic Beanstalk `aws:elasticbeanstalk:application:environment` setting | string | `DEFAULT_ENV_%d` | no |
 | env_default_value | Default ENV variable value for Elastic Beanstalk `aws:elasticbeanstalk:application:environment` setting | string | `UNSET` | no |
 | env_vars | Map of custom ENV variables to be provided to the Jenkins application running on Elastic Beanstalk, e.g. `env_vars = { JENKINS_USER = 'admin' JENKINS_PASS = 'xxxxxx' }` | map | `<map>` | no |
+| environment_type | Environment type, e.g. 'LoadBalanced' or 'SingleInstance'.  If setting to 'SingleInstance', `rolling_update_type` must be set to 'Time', `updating_min_in_service` must be set to 0, and `public_subnets` will be unused (it applies to the ELB, which does not exist in SingleInstance environments) | string | `LoadBalanced` | no |
 | force_destroy | Destroy S3 bucket for load balancer logs | string | `false` | no |
 | health_streaming_delete_on_terminate | Whether to delete the log group when the environment is terminated. If false, the health data is kept RetentionInDays days. | string | `false` | no |
 | health_streaming_enabled | For environments with enhanced health reporting enabled, whether to create a group in CloudWatch Logs for environment health and archive Elastic Beanstalk environment health data. For information about enabling enhanced health, see aws:elasticbeanstalk:healthreporting:system. | string | `false` | no |
@@ -132,7 +133,7 @@ Available targets:
 | updating_min_in_service | Minimum count of instances up during update | string | `1` | no |
 | version_label | Elastic Beanstalk Application version to deploy | string | `` | no |
 | vpc_id | ID of the VPC in which to provision the AWS resources | string | - | yes |
-| wait_for_ready_timeout | - | string | `20m` | no |
+| wait_for_ready_timeout | The maximum duration that Terraform should wait for an Elastic Beanstalk Environment to be in a ready state before timing out. | string | `20m` | no |
 | zone_id | Route53 parent zone ID. The module will create sub-domain DNS records in the parent zone for the EB environment | string | `` | no |
 
 ## Outputs

docs/terraform.md

@@ -27,6 +27,7 @@
 | env_default_key | Default ENV variable key for Elastic Beanstalk `aws:elasticbeanstalk:application:environment` setting | string | `DEFAULT_ENV_%d` | no |
 | env_default_value | Default ENV variable value for Elastic Beanstalk `aws:elasticbeanstalk:application:environment` setting | string | `UNSET` | no |
 | env_vars | Map of custom ENV variables to be provided to the Jenkins application running on Elastic Beanstalk, e.g. `env_vars = { JENKINS_USER = 'admin' JENKINS_PASS = 'xxxxxx' }` | map | `<map>` | no |
+| environment_type | Environment type, e.g. 'LoadBalanced' or 'SingleInstance'.  If setting to 'SingleInstance', `rolling_update_type` must be set to 'Time', `updating_min_in_service` must be set to 0, and `public_subnets` will be unused (it applies to the ELB, which does not exist in SingleInstance environments) | string | `LoadBalanced` | no |
 | force_destroy | Destroy S3 bucket for load balancer logs | string | `false` | no |
 | health_streaming_delete_on_terminate | Whether to delete the log group when the environment is terminated. If false, the health data is kept RetentionInDays days. | string | `false` | no |
 | health_streaming_enabled | For environments with enhanced health reporting enabled, whether to create a group in CloudWatch Logs for environment health and archive Elastic Beanstalk environment health data. For information about enabling enhanced health, see aws:elasticbeanstalk:healthreporting:system. | string | `false` | no |
@@ -69,7 +70,7 @@
 | updating_min_in_service | Minimum count of instances up during update | string | `1` | no |
 | version_label | Elastic Beanstalk Application version to deploy | string | `` | no |
 | vpc_id | ID of the VPC in which to provision the AWS resources | string | - | yes |
-| wait_for_ready_timeout | - | string | `20m` | no |
+| wait_for_ready_timeout | The maximum duration that Terraform should wait for an Elastic Beanstalk Environment to be in a ready state before timing out. | string | `20m` | no |
 | zone_id | Route53 parent zone ID. The module will create sub-domain DNS records in the parent zone for the EB environment | string | `` | no |
 
 ## Outputs

main.tf

@@ -658,6 +658,11 @@ resource "aws_elastic_beanstalk_environment" "default" {
     name      = "Application Healthcheck URL"
     value     = "HTTP:80${var.healthcheck_url}"
   }
+  setting {
+    namespace = "aws:elasticbeanstalk:environment"
+    name      = "EnvironmentType"
+    value     = "${var.environment_type}"
+  }
   setting {
     namespace = "aws:elasticbeanstalk:environment"
     name      = "LoadBalancerType"

variables.tf

@@ -100,6 +100,11 @@ variable "logs_retention_in_days" {
   description = "The number of days to keep log events before they expire."
 }
 
+variable "environment_type" {
+  default     = "LoadBalanced"
+  description = "Environment type, e.g. 'LoadBalanced' or 'SingleInstance'.  If setting to 'SingleInstance', `rolling_update_type` must be set to 'Time', `updating_min_in_service` must be set to 0, and `public_subnets` will be unused (it applies to the ELB, which does not exist in SingleInstance environments)"
+}
+
 variable "loadbalancer_type" {
   default     = "classic"
   description = "Load Balancer type, e.g. 'application' or 'classic'"
@@ -295,7 +300,8 @@ variable "solution_stack_name" {
 }
 
 variable "wait_for_ready_timeout" {
-  default = "20m"
+  default     = "20m"
+  description = "The maximum duration that Terraform should wait for an Elastic Beanstalk Environment to be in a ready state before timing out."
 }
 
 # From: http://docs.aws.amazon.com/general/latest/gr/rande.html#elasticbeanstalk_region