Google Managed Instance Group Terraform Module

Terraform module which creates managed instance groups on Google's Cloud Platform. Configuration of autoscaling can also be implemented through this module.

Usage

module my_instance_group {
  source = "garbetjie/instance-group/google"
  
  // Required parameters.
  name = "my-instance-group"  // Or name = "my-instance-group-{region}" if you want to include the region in the name.
  regions = ["europe-west4"]
  instance_template_link = google_compute_instance_template.my_instance_template.self_link
  
  // Optional parameters.
  target_size = 1
  min_replicas = 1
  max_replicas = 3
  cooldown_period = 60
  autoscaling = []
}

resource google_compute_instance_template my_instance_template {
  // ...
}

With autoscaling enabled.

It is also possible to enable autoscaling with this module:

module my_instance_group {
  source = "garbetjie/instance-group/google"
  name = "my-instance-group"  // Or name = "my-instance-group-{region}" if you want to include the region in the name.
  regions = ["europe-west4"]
  instance_template_link = google_compute_instance_template.my_instance_template.self_link
  autoscaling = [
    { cpu = 0.6 },
    { metric = "custom.googleapis.com/my_metric", value = 1, type = "GAUGE" }
  ]
}

resource google_compute_instance_template my_instance_template {
  // ...
}

Inputs

Name	Description	Type	Default	Required
name	Name format to use when creating groups. Any values of `{region}` are replaced with the instance group's region.	string	n/a	Yes
regions	Regions in which to create managed instance groups.	list(string)	n/a	Yes
instance_template_link	Link to the instance template to assign to the instance group.	string	n/a	Yes
cooldown_period	Length of time (in seconds) each new instance is allowed to boot before beginning to fetch information from it.	number	`60`	No
target_size	Number of instances in the instance group, when autoscaling is not in effect.	number	`1`	No
min_replicas	Minimum number of instances to scale down to when autoscaling is enabled.	number	`1`	No
max_replicas	Maximum number of instances to scale up to when autoscaling is enabled.	number	`3`	No
autoscaling	A list of objects that define the autoscaling to use on the instance group. Leave empty to disable autoscaling. See documentation below on the structure.	list(object)	`[]`	No

Autoscaling configuration

When enabling autoscaling, there are two "kinds" of autoscaling to use - either CPU-based, or based on the value of a Stackdriver metric. Each object in the autoscaling input parameter determines the kind of autoscaling to enable.

CPU autoscaling.

This can only be specified once in the list.

Name	Description	Type	Default	Required
cpu	The average CPU usage to maintain across the instance group.	number	n/a	Yes

Stackdriver metric autoscaling.

Name	Description	Type	Default	Required
metric	The name of the metric to autoscale on.	string	n/a	Yes
value	The value of the metric to maintain.	number	n/a	Yes
type	The type of metric being monitored. Can be one of `GAUGE`, `DELTA_PER_SECOND`, or `DELTA_PER_MINUTE`	string	`GAUGE`	No

Outputs

Name	Type	Description
manager_links	list(string)	Full URLs to all the instance group managers managed by this module.
instance_group_links	list(string)	Full URLs to all the instance groups managed this module's instance group managers.

garbetjie/terraform-google-instance-group