Changelog

v3.0.0

January 19, 2021.

Updated the Terraform AWS provider to version 3.9.0. The new provider is not backwards compatible. We had to add a field in order to provision Luster clusters.
Enabled AWS Container Insights by default for all clusters created with the containers module. Users might notice a small increase in their bill due to the increased writes to AWS CloudWatch, but it should be worth it given how much easier it makes debugging containers.
Changed the Comment field on AWS Route 53 Hosted Zones from the value "Managed by Terraform" to a more descriptive comment like “Provose private zone for module [...], DNS path [...com], and VPC vpc-[...]."
We now correctly propagate AWS resource tags from ECS services to the tasks they create. This should help users more accurately track their cloud costs per Provose module.

October 5, 2020.

Replaced the v1.x ec2_instances module with the brand-new (and slightly incompatible) ec2_on_demand_instances module.
Upgraded our pin of the Terraform AWS provider version 2.54.0 to 3.0.0. The new provider version is not backwards-compatible. We had to make some changes in how we verify Amazon Certificate Manager (ACM) certificates.
Updated the following Terraform providers: TLS to 2.2.0 and random to 2.3.0.
You can now specify a custom user for containers, just like with docker run --user.
Added the ability to specify arbitrary HTTPS redirects using the https_redirects module. This uses an AWS Application Load Balancer to route traffic from any Route 53 zone in the AWS account to any target on the web.
Added support for AWS FSx Lustre–Amazon Web Services’ managed offering of the high-performance Lustre filesystem–in the lustre_file_systems module.
Added support for AWS Elastic File Systems (EFS) by creating the elastic_file_systems module module.
Added support for AWS Batch–via the batch module.
Added support for granting access to S3 buckets for EC2 On-Demand instances (ec2_on_demand_instances) and EC2 Spot instances (ec2_spot_instances).

July 21, 2020.

Add the internal_http_health_check_success_status_codes parameter to the containers module.
Added support for AWS Fargate Spot instances to the containers module.
Specify that AWS ECS EC2 host instances should not communicate to ECS via an HTTP proxy, if the user happens to have otherwise set up a proxy on the instance.

June 28, 2020.

Various minor fixes to the Provose documentation.
We checked in an example terraform.tf to the root of the Provose GitHub repository, but this file was causing errors with Terraform so we subsequently removed it.

June 22, 2020.

Fixed an error when computing the for_each values for Elastic File System mount targets.

June 19, 2020.

This is the initial release, so from a philosophical standpoint, there have not been any changes yet.

Despite this being marked as a 1.0 release, Provose is still alpha-quality software.