Atlantis is tool which provides unified workflow for collaborating on Terraform through GitHub, GitLab and Bitbucket Cloud.
This repository contains Terraform infrastructure code which creates AWS resources required to run Atlantis on AWS, including:
- Virtual Private Cloud (VPC)
- SSL certificate using Amazon Certificate Manager (ACM)
- Application Load Balancer (ALB)
- Domain name using AWS Route53 which points to ALB
- AWS Elastic Cloud Service (ECS) and AWS Fargate running Atlantis Docker image
- AWS Parameter Store to keep secrets and access them in ECS task natively
AWS Fargate with optional support for Fargate Spot is used to reduce the bill, and it is also a cool AWS service.
Depending on which SCM system you use, Github repositories or Gitlab projects has to be configured to post events to Atlantis webhook URL.
See README.md
in examples
for Github or Gitlab for complete details.
Before using Atlantis and the code in this repository please make sure that you have read and understood the security implications described in the official Atlantis documentation.
As often with the code published in terraform-aws-modules GitHub organization you should have everything to run this code and get Atlantis up and running.
There are three ways to do this:
- Clone this github repository:
$ git clone git@github.com:terraform-aws-modules/terraform-aws-atlantis.git
$ cd terraform-aws-atlantis
-
Copy sample
terraform.tfvars.sample
intoterraform.tfvars
and specify required variables there. -
Run
terraform init
to download required providers and modules. -
Run
terraform apply
to apply the Terraform configuration and create required infrastructure. -
Run
terraform output atlantis_url
to get URL where Atlantis is publicly reachable. (Note: It may take a minute or two to get it reachable for the first time) -
Github webhook is automatically created if
github_token
,github_owner
andgithub_repo_names
were specified. Read Add GitHub Webhook in the official Atlantis documentation or check example "GitHub repository webhook for Atlantis" to add more webhooks.
This way allows integration with your existing Terraform configurations.
module "atlantis" {
source = "terraform-aws-modules/atlantis/aws"
version = "~> 3.0"
name = "atlantis"
# VPC
cidr = "10.20.0.0/16"
azs = ["eu-west-1a", "eu-west-1b", "eu-west-1c"]
private_subnets = ["10.20.1.0/24", "10.20.2.0/24", "10.20.3.0/24"]
public_subnets = ["10.20.101.0/24", "10.20.102.0/24", "10.20.103.0/24"]
# DNS (without trailing dot)
route53_zone_name = "example.com"
# ACM (SSL certificate) - Specify ARN of an existing certificate or new one will be created and validated using Route53 DNS
certificate_arn = "arn:aws:acm:eu-west-1:135367859851:certificate/70e008e1-c0e1-4c7e-9670-7bb5bd4f5a84"
# Atlantis
atlantis_github_user = "atlantis-bot"
atlantis_github_user_token = "examplegithubtoken"
atlantis_repo_allowlist = ["github.com/terraform-aws-modules/*"]
}
server-atlantis.yaml
repos:
- id: /.*/
allow_custom_workflows: true
allowed_overrides:
- apply_requirements
- workflow
apply_requirements:
- approved
workflow: default
main.tf
module "atlantis" {
source = "terraform-aws-modules/atlantis/aws"
# ...
custom_environment_variables = [
{
name : "ATLANTIS_REPO_CONFIG_JSON",
value : jsonencode(yamldecode(file("${path.module}/server-atlantis.yaml"))),
},
]
# ...
}
This way allows integration with your existing AWS resources - VPC, public and private subnets. Specify the following arguments (see methods described above):
vpc_id = "vpc-1651acf1"
private_subnet_ids = ["subnet-1fe3d837", "subnet-129d66ab"]
public_subnet_ids = ["subnet-1211eef5", "subnet-163466ab"]
If vpc_id
is specified it will take precedence over cidr
and existing VPC will be used. private_subnet_ids
and public_subnet_ids
must be specified also.
Make sure that both private and public subnets were created in the same set of availability zones (ALB will be created in public subnets, ECS Fargate service in private subnets).
If all provided subnets are public (no NAT gateway) then ecs_service_assign_public_ip
should be set to true
.
You can use service like Auth0 to secure access to Atlantis and require authentication on ALB. To enable this, you need to create Auth0 application and provide correct arguments to Atlantis module. Make sure to update application hostname, client id and client secret:
Read more in this post.
alb_authenticate_oidc = {
issuer = "https://youruser.eu.auth0.com/"
token_endpoint = "https://youruser.eu.auth0.com/oauth/token"
user_info_endpoint = "https://youruser.eu.auth0.com/userinfo"
authorization_endpoint = "https://youruser.eu.auth0.com/authorize"
authentication_request_extra_params = {}
client_id = "clientid"
client_secret = "secret123" # a data source would be good here
}
alb_authenticate_oidc = {
issuer = "https://dev-42069.okta.com/"
token_endpoint = "https://dev-42069.okta.com/oauth2/v1/token"
user_info_endpoint = "https://dev-42069.okta.com/oauth2/v1/userinfo"
authorization_endpoint = "https://dev-42069.okta.com/oauth2/v1/authorize"
authentication_request_extra_params = {}
client_id = "clientid"
client_secret = "secret123" # a data source would be good here
}
Read more in this post
alb_authenticate_oidc = {
issuer = "https://accounts.google.com"
token_endpoint = "https://oauth2.googleapis.com/token"
user_info_endpoint = "https://openidconnect.googleapis.com/v1/userinfo"
authorization_endpoint = "https://accounts.google.com/o/oauth2/v2/auth"
authentication_request_extra_params = {}
client_id = "google_client_id"
client_secret = "google_client_secret"
}
Note: remember to set your google consent screen to internal to only allow users from your own domain.
The AWS Cognito service allows you to define SAML applications tied to an identity provider (e.g., GSuite). The Atlantis ALB can then be configured to require an authenticated user managed by your identity provider.
To configure AWS Cognito connecting to a GSuite SAML application, you can use the gsuite-saml-cognito Terraform module.
To enable Cognito authentication on the Atlantis ALB, specify the following arguments containing attributes from your Cognito configuration.
alb_authenticate_cognito = {
user_pool_arn = "arn:aws:cognito-idp:us-west-2:1234567890:userpool/us-west-2_aBcDeFG"
user_pool_client_id = "clientid123"
user_pool_domain = "sso.your-corp.com"
}
If you are using one of the authentication methods above along with managed GitHub (not self-hosted enterprise version), you'll need to allow unauthenticated access to GitHub's Webhook static IPs:
allow_unauthenticated_access = true
allow_github_webhooks = true
- AWS Route53 zone is not created by this module, so zone specified as a value in
route53_zone_name
should be created before using this module. Check documentation for aws_route53_zone. - Currently this module configures Atlantis in a way that it can not be used to work with GitHub and Gitlab simultaneously (can't make list of ECS secrets conditionally).
- For Bitbucket Cloud webhook configuration follow instructions in the official Atlantis documentation.
- Complete Atlantis with GitHub webhook
- GitHub repository webhook for Atlantis
- GitLab repository webhook for Atlantis
Name | Version |
---|---|
terraform | >= 0.13.1 |
aws | ~> 3.45 |
random | >= 2.0 |
Name | Version |
---|---|
aws | ~> 3.45 |
random | >= 2.0 |
Name | Source | Version |
---|---|---|
acm | terraform-aws-modules/acm/aws | v3.2.0 |
alb | terraform-aws-modules/alb/aws | v6.5.0 |
alb_http_sg | terraform-aws-modules/security-group/aws//modules/http-80 | v4.3.0 |
alb_https_sg | terraform-aws-modules/security-group/aws//modules/https-443 | v4.3.0 |
atlantis_sg | terraform-aws-modules/security-group/aws | v4.3.0 |
container_definition_bitbucket | cloudposse/ecs-container-definition/aws | v0.58.1 |
container_definition_github_gitlab | cloudposse/ecs-container-definition/aws | v0.58.1 |
ecs | terraform-aws-modules/ecs/aws | v3.3.0 |
efs_sg | terraform-aws-modules/security-group/aws//modules/nfs | v4.8.0 |
vpc | terraform-aws-modules/vpc/aws | v3.6.0 |
Name | Description | Type | Default | Required |
---|---|---|---|---|
acm_certificate_domain_name | Route53 domain name to use for ACM certificate. Route53 zone for this domain should be created in advance. Specify if it is different from value in route53_zone_name |
string |
"" |
no |
alb_authenticate_cognito | Map of AWS Cognito authentication parameters to protect ALB (eg, using SAML). See https://www.terraform.io/docs/providers/aws/r/lb_listener.html#authenticate-cognito-action | any |
{} |
no |
alb_authenticate_oidc | Map of Authenticate OIDC parameters to protect ALB (eg, using Auth0). See https://www.terraform.io/docs/providers/aws/r/lb_listener.html#authenticate-oidc-action | any |
{} |
no |
alb_drop_invalid_header_fields | Indicates whether invalid header fields are dropped in application load balancers. Defaults to false. | bool |
null |
no |
alb_enable_deletion_protection | If true, deletion of the load balancer will be disabled via the AWS API. This will prevent Terraform from deleting the load balancer. Defaults to false. | bool |
null |
no |
alb_http_security_group_tags | Additional tags to put on the http security group | map(string) |
{} |
no |
alb_https_security_group_tags | Additional tags to put on the https security group | map(string) |
{} |
no |
alb_ingress_cidr_blocks | List of IPv4 CIDR ranges to use on all ingress rules of the ALB. | list(string) |
[ |
no |
alb_listener_ssl_policy_default | The security policy if using HTTPS externally on the load balancer. See. | string |
"ELBSecurityPolicy-2016-08" |
no |
alb_log_bucket_name | S3 bucket (externally created) for storing load balancer access logs. Required if alb_logging_enabled is true. | string |
"" |
no |
alb_log_location_prefix | S3 prefix within the log_bucket_name under which logs are stored. | string |
"" |
no |
alb_logging_enabled | Controls if the ALB will log requests to S3. | bool |
false |
no |
allow_github_webhooks | Whether to allow access for GitHub webhooks | bool |
false |
no |
allow_repo_config | When true allows the use of atlantis.yaml config files within the source repos. | string |
"false" |
no |
allow_unauthenticated_access | Whether to create ALB listener rule to allow unauthenticated access for certain CIDR blocks (eg. allow GitHub webhooks to bypass OIDC authentication) | bool |
false |
no |
allow_unauthenticated_access_priority | ALB listener rule priority for allow unauthenticated access rule | number |
10 |
no |
allow_unauthenticated_webhook_access_priority | ALB listener rule priority for allow unauthenticated webhook access rule | number |
15 |
no |
atlantis_bitbucket_base_url | Base URL of Bitbucket Server, use for Bitbucket on prem (Stash) | string |
"" |
no |
atlantis_bitbucket_user | Bitbucket username that is running the Atlantis command | string |
"" |
no |
atlantis_bitbucket_user_token | Bitbucket token of the user that is running the Atlantis command | string |
"" |
no |
atlantis_bitbucket_user_token_ssm_parameter_name | Name of SSM parameter to keep atlantis_bitbucket_user_token | string |
"/atlantis/bitbucket/user/token" |
no |
atlantis_fqdn | FQDN of Atlantis to use. Set this only to override Route53 and ALB's DNS name. | string |
null |
no |
atlantis_github_user | GitHub username that is running the Atlantis command | string |
"" |
no |
atlantis_github_user_token | GitHub token of the user that is running the Atlantis command | string |
"" |
no |
atlantis_github_user_token_ssm_parameter_name | Name of SSM parameter to keep atlantis_github_user_token | string |
"/atlantis/github/user/token" |
no |
atlantis_github_webhook_secret | GitHub webhook secret of an app that is running the Atlantis command | string |
"" |
no |
atlantis_gitlab_hostname | Gitlab server hostname, defaults to gitlab.com | string |
"gitlab.com" |
no |
atlantis_gitlab_user | Gitlab username that is running the Atlantis command | string |
"" |
no |
atlantis_gitlab_user_token | Gitlab token of the user that is running the Atlantis command | string |
"" |
no |
atlantis_gitlab_user_token_ssm_parameter_name | Name of SSM parameter to keep atlantis_gitlab_user_token | string |
"/atlantis/gitlab/user/token" |
no |
atlantis_hide_prev_plan_comments | Enables atlantis server --hide-prev-plan-comments hiding previous plan comments on update | string |
"false" |
no |
atlantis_image | Docker image to run Atlantis with. If not specified, official Atlantis image will be used | string |
"" |
no |
atlantis_log_level | Log level that Atlantis will run with. Accepted values are: <debug|info|warn|error> | string |
"debug" |
no |
atlantis_port | Local port Atlantis should be running on. Default value is most likely fine. | number |
4141 |
no |
atlantis_repo_allowlist | List of allowed repositories Atlantis can be used with | list(string) |
n/a | yes |
atlantis_security_group_tags | Additional tags to put on the atlantis security group | map(string) |
{} |
no |
atlantis_version | Verion of Atlantis to run. If not specified latest will be used | string |
"latest" |
no |
azs | A list of availability zones in the region | list(string) |
[] |
no |
certificate_arn | ARN of certificate issued by AWS ACM. If empty, a new ACM certificate will be created and validated using Route53 DNS | string |
"" |
no |
cidr | The CIDR block for the VPC which will be created if vpc_id is not specified |
string |
"" |
no |
cloudwatch_log_retention_in_days | Retention period of Atlantis CloudWatch logs | number |
7 |
no |
cloudwatch_logs_kms_key_id | The ARN of the KMS Key to use when encrypting log data. | string |
null |
no |
command | The command that is passed to the container | list(string) |
null |
no |
container_cpu | The number of cpu units used by the atlantis container. If not specified ecs_task_cpu will be used | number |
null |
no |
container_depends_on | The dependencies defined for container startup and shutdown. A container can contain multiple dependencies. When a dependency is defined for container startup, for container shutdown it is reversed. The condition can be one of START, COMPLETE, SUCCESS or HEALTHY | list(object({ |
null |
no |
container_memory | The amount (in MiB) of memory used by the atlantis container. If not specified ecs_task_memory will be used | number |
null |
no |
container_memory_reservation | The amount of memory (in MiB) to reserve for the container | number |
128 |
no |
create_ecs_cluster | Whether to create an ECS cluster or not | bool |
true |
no |
create_route53_record | Whether to create Route53 record for Atlantis | bool |
true |
no |
custom_container_definitions | A list of valid container definitions provided as a single valid JSON document. By default, the standard container definition is used. | string |
"" |
no |
custom_environment_secrets | List of additional secrets the container will use (list should contain maps with name and valueFrom ) |
list(object( |
[] |
no |
custom_environment_variables | List of additional environment variables the container will use (list should contain maps with name and value ) |
list(object( |
[] |
no |
default_security_group_egress | List of maps of egress rules to set on the default security group | list(map(string)) |
[] |
no |
default_security_group_ingress | List of maps of ingress rules to set on the default security group | list(map(string)) |
[] |
no |
docker_labels | The configuration options to send to the docker_labels |
map(string) |
null |
no |
ecs_cluster_id | ID of an existing ECS cluster where resources will be created | string |
"" |
no |
ecs_container_insights | Controls if ECS Cluster has container insights enabled | bool |
false |
no |
ecs_fargate_spot | Whether to run ECS Fargate Spot or not | bool |
false |
no |
ecs_service_assign_public_ip | Should be true, if ECS service is using public subnets (more info: https://docs.aws.amazon.com/AmazonECS/latest/developerguide/task_cannot_pull_image.html) | bool |
false |
no |
ecs_service_deployment_maximum_percent | The upper limit (as a percentage of the service's desiredCount) of the number of running tasks that can be running in a service during a deployment | number |
100 |
no |
ecs_service_deployment_minimum_healthy_percent | The lower limit (as a percentage of the service's desiredCount) of the number of running tasks that must remain running and healthy in a service during a deployment | number |
0 |
no |
ecs_service_desired_count | The number of instances of the task definition to place and keep running | number |
1 |
no |
ecs_service_enable_execute_command | Enable ECS exec for the service. This can be used to allow interactive sessions and commands to be executed in the container | bool |
true |
no |
ecs_service_force_new_deployment | Enable to force a new task deployment of the service. This can be used to update tasks to use a newer Docker image with same image/tag combination (e.g. myimage:latest) | bool |
false |
no |
ecs_service_platform_version | The platform version on which to run your service | string |
"LATEST" |
no |
ecs_task_cpu | The number of cpu units used by the task | number |
256 |
no |
ecs_task_memory | The amount (in MiB) of memory used by the task | number |
512 |
no |
enable_ecs_managed_tags | Specifies whether to enable Amazon ECS managed tags for the tasks within the service | bool |
false |
no |
enable_ephemeral_storage | Enable to use Fargate Ephemeral Storage | bool |
false |
no |
entrypoint | The entry point that is passed to the container | list(string) |
null |
no |
ephemeral_storage_size | Size of Ephemeral Storage in GiB | number |
21 |
no |
essential | Determines whether all other containers in a task are stopped, if this container fails or stops for any reason. Due to how Terraform type casts booleans in json it is required to double quote this value | bool |
true |
no |
external_task_definition_updates | Enable to allow the task definition to be updated outside of this Terraform module. This should be enabled when using a deployment tool such as ecs-deploy which updates the task definition and will then keep the ECS service using the latest version of the task definition. | bool |
false |
no |
extra_container_definitions | A list of valid container definitions provided as a single valid JSON document. These will be provided as supplimentary to the main Atlantis container definition | list(any) |
[] |
no |
extra_load_balancers | A list of maps for additional ECS task load balancers | list(map(string)) |
[] |
no |
firelens_configuration | The FireLens configuration for the container. This is used to specify and configure a log router for container logs. For more details, see https://docs.aws.amazon.com/AmazonECS/latest/APIReference/API_FirelensConfiguration.html | object({ |
null |
no |
github_webhooks_cidr_blocks | List of CIDR blocks used by GitHub webhooks | list(string) |
[ |
no |
internal | Whether the load balancer is internal or external | bool |
false |
no |
manage_default_security_group | Should be true to adopt and manage default security group | bool |
false |
no |
mount_points | Container mount points. This is a list of maps, where each map should contain a containerPath and sourceVolume . The readOnly key is optional. |
list(any) |
[] |
no |
name | Name to use on all resources created (VPC, ALB, etc) | string |
"atlantis" |
no |
permissions_boundary | If provided, all IAM roles will be created with this permissions boundary attached. | string |
null |
no |
policies_arn | A list of the ARN of the policies you want to apply | list(string) |
null |
no |
private_subnet_ids | A list of IDs of existing private subnets inside the VPC | list(string) |
[] |
no |
private_subnets | A list of private subnets inside the VPC | list(string) |
[] |
no |
propagate_tags | Specifies whether to propagate the tags from the task definition or the service to the tasks. The valid values are SERVICE and TASK_DEFINITION | string |
null |
no |
public_subnet_ids | A list of IDs of existing public subnets inside the VPC | list(string) |
[] |
no |
public_subnets | A list of public subnets inside the VPC | list(string) |
[] |
no |
readonly_root_filesystem | Determines whether a container is given read-only access to its root filesystem. Due to how Terraform type casts booleans in json it is required to double quote this value | bool |
false |
no |
repository_credentials | Container repository credentials; required when using a private repo. This map currently supports a single key; "credentialsParameter", which should be the ARN of a Secrets Manager's secret holding the credentials | map(string) |
null |
no |
route53_private_zone | Enable to use a private Route53 zone | bool |
false |
no |
route53_record_name | Name of Route53 record to create ACM certificate in and main A-record. If null is specified, var.name is used instead. Provide empty string to point root domain name to ALB. | string |
null |
no |
route53_zone_name | Route53 zone name to create ACM certificate in and main A-record, without trailing dot | string |
"" |
no |
security_group_ids | List of one or more security groups to be added to the load balancer | list(string) |
[] |
no |
ssm_kms_key_arn | ARN of KMS key to use for encryption and decryption of SSM Parameters. Required only if your key uses a custom KMS key and not the default key | string |
"" |
no |
start_timeout | Time duration (in seconds) to wait before giving up on resolving dependencies for a container | number |
30 |
no |
stop_timeout | Time duration (in seconds) to wait before the container is forcefully killed if it doesn't exit normally on its own | number |
30 |
no |
tags | A map of tags to use on all resources | map(string) |
{} |
no |
trusted_entities | A list of users or roles, that can assume the task role | list(string) |
[] |
no |
trusted_principals | A list of principals, in addition to ecs-tasks.amazonaws.com, that can assume the task role | list(string) |
[] |
no |
ulimits | Container ulimit settings. This is a list of maps, where each map should contain "name", "hardLimit" and "softLimit" | list(object({ |
null |
no |
use_ecs_old_arn_format | A flag to enable/disable tagging the ecs resources that require the new longer arn format | bool |
false |
no |
user | The user to run as inside the container. Must be in the uid:gid or the default (null) will use the container's configured USER directive or root if not set. |
string |
null |
no |
volumes_from | A list of VolumesFrom maps which contain "sourceContainer" (name of the container that has the volumes to mount) and "readOnly" (whether the container can write to the volume) | list(object({ |
[] |
no |
vpc_id | ID of an existing VPC where resources will be created | string |
"" |
no |
webhook_ssm_parameter_name | Name of SSM parameter to keep webhook secret | string |
"/atlantis/webhook/secret" |
no |
whitelist_unauthenticated_cidr_blocks | List of allowed CIDR blocks to bypass authentication | list(string) |
[] |
no |
working_directory | The working directory to run commands inside the container | string |
null |
no |
Name | Description |
---|---|
alb_arn | ARN of alb |
alb_dns_name | Dns name of alb |
alb_http_listeners_arn | ARNs of alb http listeners |
alb_http_listeners_id | Ids of alb http listeners |
alb_https_listeners_arn | ARN of alb https listeners |
alb_https_listeners_id | Ids of alb https listeners |
alb_security_group_id | Security group of alb |
alb_zone_id | Zone ID of alb |
atlantis_repo_allowlist | Git repositories where webhook should be created |
atlantis_url | URL of Atlantis |
atlantis_url_events | Webhook events URL of Atlantis |
ecs_cluster_arn | ECS cluster ARN |
ecs_cluster_id | ECS cluster id |
ecs_security_group | Security group assigned to ECS Service in network configuration |
ecs_task_definition | Task definition for ECS service (used for external triggers) |
private_subnet_ids | IDs of the VPC private subnets that were created or passed in |
public_subnet_ids | IDs of the VPC public subnets that were created or passed in |
task_role_arn | The Atlantis ECS task role arn |
task_role_id | The Atlantis ECS task role id |
task_role_name | The Atlantis ECS task role name |
task_role_unique_id | The stable and unique string identifying the Atlantis ECS task role. |
vpc_id | ID of the VPC that was created or passed in |
webhook_secret | Webhook secret |
Module is maintained by Anton Babenko with help from these awesome contributors.
Apache 2 Licensed. See LICENSE for full details.