terraform-aws-autoscaling

Description

A terraform module which provisions an auto scaling group along with its launch configuration.

Table of Content

• terraform-aws-autoscaling
  • Description
  • Dependencies
  • Getting Started
    • Conventions
    • Behaviour
    • Migration from pre-launch-template versions
    • Switching between plain launch template and mixed instance policy
  • Examples
  • Terraform Version
  • Requirements
  • Providers
  • Modules
  • Resources
  • Inputs
  • Outputs
  • Authors
  • Contributing
  • License

Dependencies

This Terraform module uses another Terraform module, here is the list of Terraform module dependencies:

• traveloka/terraform-aws-resource-naming

Getting Started

Conventions

• the auto scaling group will have Service, Cluster, Environment, and ProductDomain tags by default, which are propagated to all instances it spawns

Behaviour

• To specify on-demand instance types, use the launch_template_overrides variable. Auto Scaling will launch instances based on the order of preference specified in that list. ["c5.large", "c4.large", "m5.large"] means the ASG will always try to launch c5.large if it's available, falling back to c4.large if it's not available, and falling back to m5.large if the previous two aren't available
• On the first deployment, this module will provision an ASG with a launch template that select the most recent AMI that passes through the given image_filters
• Each time there's a change in the values of the module.asg_name's keepers (e.g. security group, AMI ID), a new ASG will be provisioned by terraform, and the old one will later be destroyed (doing the "simple swap" deployment strategy).
• When there's a change in launch template parameters' values, terraform will create a new launch template version unless the new configuration is already the same as the latest version of the launch template (e.g. when the launch template had been updated externally).

Migration from pre-launch-template versions

terraform init
terraform state rm module.<this module name in your terraform code>.module.random_lc
terraform apply

Switching between plain launch template and mixed instance policy

switching to plain launch template

module "asg" {
  source = "github.com/traveloka/terraform-aws-autoscaling?ref=v0.3.1"
  # ...
  use_mixed_instance_policy = false
  launch_template_overrides = [
    {
      "instance_type" = "c4.large" # this (the first element) will be the launch template's instance type
    },
    {
      "instance_type" = "t3.medium"
    },
  ]
}

switching to mixed instance policy

module "asg" {
  source = "github.com/traveloka/terraform-aws-autoscaling?ref=v0.3.1"
  # ...
  use_mixed_instance_policy = true
}

Examples

• Simple
• Public

Terraform Version

This module was created on 15/08/2018. The latest stable version of Terraform which this module tested working is Terraform 0.13.7 on 28/09/2021.

Requirements

Name	Version
terraform	>= 0.13

Providers

Name	Version
aws	n/a

Modules

Name	Source	Version
asg_name	github.com/traveloka/terraform-aws-resource-naming.git	v0.20.0
launch_template_name	github.com/traveloka/terraform-aws-resource-naming.git	v0.20.0

Resources

Name	Type
aws_autoscaling_group.main	resource
aws_launch_template.main	resource
aws_ami.latest_service_image	data source

Inputs

Name	Description	Type	Default	Required
application	Application type that the ASG's instances will serve	string	n/a	yes
asg_clb_names	A list of classic load balancer names to add to the autoscaling group	list(string)	[]	no
asg_default_cooldown	Time, in seconds, the minimum interval of two scaling activities	string	"300"	no
asg_desired_capacity	The created ASG will have this number of instances at desired	string	"2"	no
asg_enabled_metrics	The list of ASG metrics to collect	list(string)	[ "GroupMinSize", "GroupMaxSize", "GroupDesiredCapacity", "GroupInServiceInstances", "GroupPendingInstances", "GroupStandbyInstances", "GroupTerminatingInstances", "GroupTotalInstances" ]	no
asg_health_check_grace_period	Time, in seconds, to wait for new instances before checking their health	string	"300"	no
asg_health_check_type	Controls how ASG health checking is done	string	"ELB"	no
asg_lb_target_group_arns	The created ASG will be attached to this target group	list(string)	[]	no
asg_max_capacity	The created ASG will have this number of instances at maximum	string	"0"	no
asg_metrics_granularity	The granularity to associate with the metrics to collect	string	"1Minute"	no
asg_min_capacity	The created ASG will have this number of instances at minimum	string	"0"	no
asg_placement_group	The placement group for the spawned instances	string	""	no
asg_service_linked_role_arn	The ARN of the service-linked role that the ASG will use to call other AWS services	string	""	no
asg_tags	The created ASG will have these tags applied over the default ones (see main.tf)	list(map(string))	[]	no
asg_termination_policies	Specify policies that the auto scaling group should use to terminate its instances	list(string)	[ "Default" ]	no
asg_vpc_zone_identifier	The created ASG will spawn instances to these subnet IDs	list(string)	n/a	yes
asg_wait_for_capacity_timeout	A maximum duration that Terraform should wait for ASG instances to be healthy before timing out	string	n/a	yes
asg_wait_for_elb_capacity	Terraform will wait for exactly this number of healthy instances in all attached load balancers on both create and update operations. If left to default, the value is set to asg_min_capacity	string	""	no
associate_public_ip	Whether to associate public IP to the instance	string	"false"	no
cluster_role	Primary role/function of the cluster	string	"app"	no
cpu_credits	The credit option for CPU usage, can be either 'standard' or 'unlimited'	string	"unlimited"	no
delete_network_interface_on_termination	Whether the network interface will be deleted on termination	string	"true"	no
delete_on_termination	Whether the volume should be destroyed on instance termination	string	"true"	no
description	Free form description of this ASG and its instances	string	n/a	yes
disable_api_termination	whether to protect your instance from accidently being terminated from console or api	bool	false	no
ebs_encryption	Whether the volume will be encrypted or not	string	"false"	no
ebs_optimized	The spawned instances will have EBS optimization if enabled	string	"false"	no
environment	The created resources will belong to this infrastructure environment	string	n/a	yes
image_filters	The AMI search filters. The most recent AMI that pass this filter will be deployed to the ASG. See https://docs.aws.amazon.com/cli/latest/reference/ec2/describe-images.html	list(object({ name = string, values = list(string) }) )	n/a	yes
image_owners	List of AMI owners to limit search. This becomes required starting terraform-provider-aws v2 (https://www.terraform.io/docs/providers/aws/guides/version-2-upgrade.html#owners-argument-now-required)	list(string)	n/a	yes
instance_profile_name	The spawned instances will have this IAM profile	string	n/a	yes
key_name	The spawned instances will have this SSH key name	string	""	no
launch_template_overrides	List of nested arguments provides the ability to specify multiple instance types. See https://www.terraform.io/docs/providers/aws/r/autoscaling_group.html#override When using plain launch template, the first element's instance_type will be used as the launch template instance type.	list(map(string))	[ { "instance_type": "c4.large" }, { "instance_type": "t3.medium" } ]	no
mixed_instances_distribution	Specify the distribution of on-demand instances and spot instances. See https://docs.aws.amazon.com/autoscaling/ec2/APIReference/API_InstancesDistribution.html	map(string)	{ "on_demand_allocation_strategy": "prioritized", "on_demand_base_capacity": "0", "on_demand_percentage_above_base_capacity": "100", "spot_allocation_strategy": "lowest-price", "spot_instance_pools": "2", "spot_max_price": "" }	no
monitoring	The spawned instances will have enhanced monitoring if enabled	string	"true"	no
product_domain	Abbreviation of the product domain this ASG and its instances belongs to	string	n/a	yes
security_groups	The spawned instances will have these security groups	list(string)	n/a	yes
service_name	The name of the service	string	n/a	yes
use_mixed_instances_policy	Whether to use ASG mixed instances policy or the plain launch template	bool	true	no
user_data	The spawned instances will have this user data. Use the rendered value of a terraform's template_cloudinit_config data	string	" "	no
volume_size	The size of the volume in gigabytes	string	"8"	no
volume_type	The type of volume. Can be standard, gp2, or io1	string	"gp2"	no

Outputs

Name	Description
asg_desired_capacity	The desired capacity of the auto scaling group; it may be useful when doing blue/green asg deployment (create a new asg while copying the old's capacity)
asg_name	The name of the auto scaling group
launch_template_name	The name of the launch template used by the auto scaling group
resource_naming_launch_template	The name of the launch template generated by resource-naming module.

Authors

• Salvian Reynaldi

Contributing

This module accepting or open for any contributions from anyone, please see the CONTRIBUTING.md for more detail about how to contribute to this module.

License

This module is under Apache License 2.0 - see the LICENSE file for details.