Terraform module to provision AWS resources to run EC2 worker nodes for Elastic Container Service for Kubernetes.
Instantiate it multiple times to create many EKS worker node pools with specific settings such as GPUs, EC2 instance types, or autoscale parameters.
This project is part of our comprehensive "SweetOps" approach towards DevOps.
It's 100% Open Source and licensed under the APACHE2.
We literally have hundreds of terraform modules that are Open Source and well-maintained. Check them out!
The module provisions the following resources:
- IAM Role and Instance Profile to allow Kubernetes nodes to access other AWS services
- Security Group with rules for EKS workers to allow networking traffic
- AutoScaling Group with Launch Template to configure and launch worker instances
- AutoScaling Policies and CloudWatch Metric Alarms to monitor CPU utilization on the EC2 instances and scale the number of instance in the AutoScaling Group up or down.
If you don't want to use the provided functionality, or want to provide your own policies, disable it by setting the variable
autoscaling_policies_enabled
to"false"
.
IMPORTANT: The master
branch is used in source
just as an example. In your code, do not pin to master
because there may be breaking changes between releases.
Instead pin to the release tag (e.g. ?ref=tags/x.y.z
) of one of our latest releases.
For a complete example, see examples/complete
provider "aws" {
region = var.region
}
locals {
# The usage of the specific kubernetes.io/cluster/* resource tags below are required
# for EKS and Kubernetes to discover and manage networking resources
# https://www.terraform.io/docs/providers/aws/guides/eks-getting-started.html#base-vpc-networking
tags = merge(var.tags, map("kubernetes.io/cluster/${var.cluster_name}", "shared"))
}
module "vpc" {
source = "git::https://github.com/cloudposse/terraform-aws-vpc.git?ref=tags/0.8.0"
namespace = var.namespace
stage = var.stage
name = var.name
cidr_block = "172.16.0.0/16"
tags = local.tags
}
module "subnets" {
source = "git::https://github.com/cloudposse/terraform-aws-dynamic-subnets.git?ref=tags/0.16.0"
availability_zones = var.availability_zones
namespace = var.namespace
stage = var.stage
name = var.name
vpc_id = module.vpc.vpc_id
igw_id = module.vpc.igw_id
cidr_block = module.vpc.vpc_cidr_block
nat_gateway_enabled = false
nat_instance_enabled = false
tags = local.tags
}
module "eks_workers" {
source = "git::https://github.com/cloudposse/terraform-aws-eks-workers.git?ref=master"
namespace = var.namespace
stage = var.stage
name = var.name
instance_type = var.instance_type
vpc_id = module.vpc.vpc_id
subnet_ids = module.subnets.public_subnet_ids
health_check_type = var.health_check_type
min_size = var.min_size
max_size = var.max_size
wait_for_capacity_timeout = var.wait_for_capacity_timeout
cluster_name = var.cluster_name
cluster_endpoint = var.cluster_endpoint
cluster_certificate_authority_data = var.cluster_certificate_authority_data
cluster_security_group_id = var.cluster_security_group_id
# Auto-scaling policies and CloudWatch metric alarms
autoscaling_policies_enabled = var.autoscaling_policies_enabled
cpu_utilization_high_threshold_percent = var.cpu_utilization_high_threshold_percent
cpu_utilization_low_threshold_percent = var.cpu_utilization_low_threshold_percent
}
Available targets:
help Help screen
help/all Display help for all targets
help/short This help short screen
lint Lint terraform code
Name | Description | Type | Default | Required |
---|---|---|---|---|
additional_security_group_ids | Additional list of security groups that will be attached to the autoscaling group | list(string) | <list> |
no |
after_cluster_joining_userdata | Additional commands to execute on each worker node after joining the EKS cluster (after executing the bootstrap.sh script). For mot info, see https://kubedex.com/90-days-of-aws-eks-in-production |
string | `` | no |
allowed_cidr_blocks | List of CIDR blocks to be allowed to connect to the worker nodes | list(string) | <list> |
no |
allowed_security_groups | List of Security Group IDs to be allowed to connect to the worker nodes | list(string) | <list> |
no |
associate_public_ip_address | Associate a public IP address with an instance in a VPC | bool | false |
no |
attributes | Additional attributes (e.g. 1 ) |
list(string) | <list> |
no |
autoscaling_policies_enabled | Whether to create aws_autoscaling_policy and aws_cloudwatch_metric_alarm resources to control Auto Scaling |
bool | true |
no |
aws_iam_instance_profile_name | The name of the existing instance profile that will be used in autoscaling group for EKS workers. If empty will create a new instance profile. | string | `` | no |
before_cluster_joining_userdata | Additional commands to execute on each worker node before joining the EKS cluster (before executing the bootstrap.sh script). For mot info, see https://kubedex.com/90-days-of-aws-eks-in-production |
string | `` | no |
block_device_mappings | Specify volumes to attach to the instance besides the volumes specified by the AMI | object | <list> |
no |
bootstrap_extra_args | Passed to the bootstrap.sh script to enable --kublet-extra-args or --use-max-pods |
string | `` | no |
cluster_certificate_authority_data | The base64 encoded certificate data required to communicate with the cluster | string | - | yes |
cluster_endpoint | EKS cluster endpoint | string | - | yes |
cluster_name | The name of the EKS cluster | string | - | yes |
cluster_security_group_id | Security Group ID of the EKS cluster | string | - | yes |
cluster_security_group_ingress_enabled | Whether to enable the EKS cluster Security Group as ingress to workers Security Group | bool | true |
no |
cpu_utilization_high_evaluation_periods | The number of periods over which data is compared to the specified threshold | number | 2 |
no |
cpu_utilization_high_period_seconds | The period in seconds over which the specified statistic is applied | number | 300 |
no |
cpu_utilization_high_statistic | The statistic to apply to the alarm's associated metric. Either of the following is supported: SampleCount , Average , Sum , Minimum , Maximum |
string | Average |
no |
cpu_utilization_high_threshold_percent | The value against which the specified statistic is compared | number | 90 |
no |
cpu_utilization_low_evaluation_periods | The number of periods over which data is compared to the specified threshold | number | 2 |
no |
cpu_utilization_low_period_seconds | The period in seconds over which the specified statistic is applied | number | 300 |
no |
cpu_utilization_low_statistic | The statistic to apply to the alarm's associated metric. Either of the following is supported: SampleCount , Average , Sum , Minimum , Maximum |
string | Average |
no |
cpu_utilization_low_threshold_percent | The value against which the specified statistic is compared | number | 10 |
no |
credit_specification | Customize the credit specification of the instances | object | null |
no |
default_cooldown | The amount of time, in seconds, after a scaling activity completes before another scaling activity can start | number | 300 |
no |
delimiter | Delimiter to be used between namespace , stage , name and attributes |
string | - |
no |
disable_api_termination | If true , enables EC2 Instance Termination Protection |
bool | false |
no |
ebs_optimized | If true, the launched EC2 instance will be EBS-optimized | bool | false |
no |
eks_worker_ami_name_filter | AMI name filter to lookup the most recent EKS AMI if image_id is not provided |
string | amazon-eks-node-* |
no |
eks_worker_ami_name_regex | A regex string to apply to the AMI list returned by AWS | string | ^amazon-eks-node-[1-9,\.]+-v\d{8}$ |
no |
elastic_gpu_specifications | Specifications of Elastic GPU to attach to the instances | object | null |
no |
enable_monitoring | Enable/disable detailed monitoring | bool | true |
no |
enabled | Whether to create the resources. Set to false to prevent the module from creating any resources |
bool | true |
no |
enabled_metrics | A list of metrics to collect. The allowed values are GroupMinSize , GroupMaxSize , GroupDesiredCapacity , GroupInServiceInstances , GroupPendingInstances , GroupStandbyInstances , GroupTerminatingInstances , GroupTotalInstances |
list(string) | <list> |
no |
force_delete | Allows deleting the autoscaling group without waiting for all instances in the pool to terminate. You can force an autoscaling group to delete even if it's in the process of scaling a resource. Normally, Terraform drains all the instances before deleting the group. This bypasses that behavior and potentially leaves resources dangling | bool | false |
no |
health_check_grace_period | Time (in seconds) after instance comes into service before checking health | number | 300 |
no |
health_check_type | Controls how health checking is done. Valid values are EC2 or ELB |
string | EC2 |
no |
image_id | EC2 image ID to launch. If not provided, the module will lookup the most recent EKS AMI. See https://docs.aws.amazon.com/eks/latest/userguide/eks-optimized-ami.html for more details on EKS-optimized images | string | `` | no |
instance_initiated_shutdown_behavior | Shutdown behavior for the instances. Can be stop or terminate |
string | terminate |
no |
instance_market_options | The market (purchasing) option for the instances | object | null |
no |
instance_type | Instance type to launch | string | - | yes |
key_name | SSH key name that should be used for the instance | string | `` | no |
load_balancers | A list of elastic load balancer names to add to the autoscaling group. Only valid for classic load balancers. For ALBs, use target_group_arns instead |
list(string) | <list> |
no |
max_size | The maximum size of the autoscale group | number | - | yes |
metrics_granularity | The granularity to associate with the metrics to collect. The only valid value is 1Minute | string | 1Minute |
no |
min_elb_capacity | Setting this causes Terraform to wait for this number of instances to show up healthy in the ELB only on creation. Updates will not wait on ELB instance number changes | number | 0 |
no |
min_size | The minimum size of the autoscale group | number | - | yes |
name | Solution name, e.g. 'app' or 'cluster' | string | app |
no |
namespace | Namespace, which could be your organization name, e.g. 'eg' or 'cp' | string | `` | no |
placement | The placement specifications of the instances | object | null |
no |
placement_group | The name of the placement group into which you'll launch your instances, if any | string | `` | no |
protect_from_scale_in | Allows setting instance protection. The autoscaling group will not select instances with this setting for terminination during scale in events | bool | false |
no |
scale_down_adjustment_type | Specifies whether the adjustment is an absolute number or a percentage of the current capacity. Valid values are ChangeInCapacity , ExactCapacity and PercentChangeInCapacity |
string | ChangeInCapacity |
no |
scale_down_cooldown_seconds | The amount of time, in seconds, after a scaling activity completes and before the next scaling activity can start | number | 300 |
no |
scale_down_policy_type | The scalling policy type, either SimpleScaling , StepScaling or TargetTrackingScaling |
string | SimpleScaling |
no |
scale_down_scaling_adjustment | The number of instances by which to scale. scale_down_scaling_adjustment determines the interpretation of this number (e.g. as an absolute number or as a percentage of the existing Auto Scaling group size). A positive increment adds to the current capacity and a negative value removes from the current capacity |
number | -1 |
no |
scale_up_adjustment_type | Specifies whether the adjustment is an absolute number or a percentage of the current capacity. Valid values are ChangeInCapacity , ExactCapacity and PercentChangeInCapacity |
string | ChangeInCapacity |
no |
scale_up_cooldown_seconds | The amount of time, in seconds, after a scaling activity completes and before the next scaling activity can start | number | 300 |
no |
scale_up_policy_type | The scalling policy type, either SimpleScaling , StepScaling or TargetTrackingScaling |
string | SimpleScaling |
no |
scale_up_scaling_adjustment | The number of instances by which to scale. scale_up_adjustment_type determines the interpretation of this number (e.g. as an absolute number or as a percentage of the existing Auto Scaling group size). A positive increment adds to the current capacity and a negative value removes from the current capacity |
number | 1 |
no |
service_linked_role_arn | The ARN of the service-linked role that the ASG will use to call other AWS services | string | `` | no |
stage | Stage, e.g. 'prod', 'staging', 'dev', or 'test' | string | `` | no |
subnet_ids | A list of subnet IDs to launch resources in | list(string) | - | yes |
suspended_processes | A list of processes to suspend for the AutoScaling Group. The allowed values are Launch , Terminate , HealthCheck , ReplaceUnhealthy , AZRebalance , AlarmNotification , ScheduledActions , AddToLoadBalancer . Note that if you suspend either the Launch or Terminate process types, it can prevent your autoscaling group from functioning properly. |
list(string) | <list> |
no |
tags | Additional tags (e.g. { BusinessUnit = "XYZ" } |
map(string) | <map> |
no |
target_group_arns | A list of aws_alb_target_group ARNs, for use with Application Load Balancing | list(string) | <list> |
no |
termination_policies | A list of policies to decide how the instances in the auto scale group should be terminated. The allowed values are OldestInstance , NewestInstance , OldestLaunchConfiguration , ClosestToNextInstanceHour , Default |
list(string) | <list> |
no |
use_custom_image_id | If set to true , will use variable image_id for the EKS workers inside autoscaling group |
bool | false |
no |
use_existing_aws_iam_instance_profile | If set to true , will use variable aws_iam_instance_profile_name to run EKS workers using an existing AWS instance profile that was created outside of this module, workaround for error like count cannot be computed |
bool | false |
no |
use_existing_security_group | If set to true , will use variable workers_security_group_id to run EKS workers using an existing security group that was created outside of this module, workaround for errors like count cannot be computed |
bool | false |
no |
vpc_id | VPC ID for the EKS cluster | string | - | yes |
wait_for_capacity_timeout | A maximum duration that Terraform should wait for ASG instances to be healthy before timing out. Setting this to '0' causes Terraform to skip all Capacity Waiting behavior | string | 10m |
no |
wait_for_elb_capacity | Setting this will cause Terraform to wait for exactly this number of healthy instances in all attached load balancers on both create and update operations. Takes precedence over min_elb_capacity behavior |
number | 0 |
no |
workers_role_policy_arns | List of policy ARNs that will be attached to the workers default role on creation | list(string) | <list> |
no |
workers_role_policy_arns_count | Count of policy ARNs that will be attached to the workers default role on creation. Needed to prevent Terraform error count can't be computed |
number | 0 |
no |
workers_security_group_id | The name of the existing security group that will be used in autoscaling group for EKS workers. If empty, a new security group will be created | string | `` | no |
Name | Description |
---|---|
autoscaling_group_arn | ARN of the AutoScaling Group |
autoscaling_group_default_cooldown | Time between a scaling activity and the succeeding scaling activity |
autoscaling_group_desired_capacity | The number of Amazon EC2 instances that should be running in the group |
autoscaling_group_health_check_grace_period | Time after instance comes into service before checking health |
autoscaling_group_health_check_type | EC2 or ELB . Controls how health checking is done |
autoscaling_group_id | The AutoScaling Group ID |
autoscaling_group_max_size | The maximum size of the AutoScaling Group |
autoscaling_group_min_size | The minimum size of the AutoScaling Group |
autoscaling_group_name | The AutoScaling Group name |
launch_template_arn | ARN of the launch template |
launch_template_id | The ID of the launch template |
security_group_arn | ARN of the worker nodes Security Group |
security_group_id | ID of the worker nodes Security Group |
security_group_name | Name of the worker nodes Security Group |
workers_role_arn | ARN of the worker nodes IAM role |
workers_role_name | Name of the worker nodes IAM role |
Like this project? Please give it a ★ on our GitHub! (it helps us a lot)
Are you using this project or any of our other projects? Consider leaving a testimonial. =)
Check out these related projects.
- terraform-aws-ec2-autoscale-group - Terraform module to provision Auto Scaling Group and Launch Template on AWS
- terraform-aws-ecs-container-definition - Terraform module to generate well-formed JSON documents (container definitions) that are passed to the aws_ecs_task_definition Terraform resource
- terraform-aws-ecs-alb-service-task - Terraform module which implements an ECS service which exposes a web service via ALB
- terraform-aws-ecs-web-app - Terraform module that implements a web app on ECS and supports autoscaling, CI/CD, monitoring, ALB integration, and much more
- terraform-aws-ecs-codepipeline - Terraform module for CI/CD with AWS Code Pipeline and Code Build for ECS
- terraform-aws-ecs-cloudwatch-autoscaling - Terraform module to autoscale ECS Service based on CloudWatch metrics
- terraform-aws-ecs-cloudwatch-sns-alarms - Terraform module to create CloudWatch Alarms on ECS Service level metrics
- terraform-aws-ec2-instance - Terraform module for providing a general purpose EC2 instance
- terraform-aws-ec2-instance-group - Terraform module for provisioning multiple general purpose EC2 hosts for stateful applications
Got a question?
File a GitHub issue, send us an email or join our Slack Community.
Work directly with our team of DevOps experts via email, slack, and video conferencing.
We provide commercial support for all of our Open Source projects. As a Dedicated Support customer, you have access to our team of subject matter experts at a fraction of the cost of a full-time engineer.
- Questions. We'll use a Shared Slack channel between your team and ours.
- Troubleshooting. We'll help you triage why things aren't working.
- Code Reviews. We'll review your Pull Requests and provide constructive feedback.
- Bug Fixes. We'll rapidly work to fix any bugs in our projects.
- Build New Terraform Modules. We'll develop original modules to provision infrastructure.
- Cloud Architecture. We'll assist with your cloud strategy and design.
- Implementation. We'll provide hands-on support to implement our reference architectures.
Are you interested in custom Terraform module development? Submit your inquiry using our form today and we'll get back to you ASAP.
Join our Open Source Community on Slack. It's FREE for everyone! Our "SweetOps" community is where you get to talk with others who share a similar vision for how to rollout and manage infrastructure. This is the best place to talk shop, ask questions, solicit feedback, and work together as a community to build totally sweet infrastructure.
Signup for our newsletter that covers everything on our technology radar. Receive updates on what we're up to on GitHub as well as awesome new projects we discover.
Please use the issue tracker to report any bugs or file feature requests.
If you are interested in being a contributor and want to get involved in developing this project or help out with our other projects, we would love to hear from you! Shoot us an email.
In general, PRs are welcome. We follow the typical "fork-and-pull" Git workflow.
- Fork the repo on GitHub
- Clone the project to your own machine
- Commit changes to your own branch
- Push your work back up to your fork
- Submit a Pull Request so that we can review your changes
NOTE: Be sure to merge the latest changes from "upstream" before making a pull request!
Copyright © 2017-2019 Cloud Posse, LLC
See LICENSE for full details.
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you under the Apache License, Version 2.0 (the
"License"); you may not use this file except in compliance
with the License. You may obtain a copy of the License at
https://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.
All other trademarks referenced herein are the property of their respective owners.
This project is maintained and funded by Cloud Posse, LLC. Like it? Please let us know by leaving a testimonial!
We're a DevOps Professional Services company based in Los Angeles, CA. We ❤️ Open Source Software.
We offer paid support on all of our projects.
Check out our other projects, follow us on twitter, apply for a job, or hire us to help with your cloud strategy and implementation.
Erik Osterman |
Andriy Knysh |
Igor Rodionov |
---|