This repository has been archived by the owner on Apr 27, 2020. It is now read-only.
Releases: cmdlabs/terraform-aws-eks
Releases · cmdlabs/terraform-aws-eks
0.8.0
0.7.0
Breaking
- KIAM no longer has dedicated nodes. This is possible now that KIAM has an option to not remove the iptables rule on shutdown resulting in a fail secure design.
- ALB Ingress no longer has security groups created and managed by this module.
Added
- Autotagging now has the capability to specify which AWS profile the aws-cli uses.
0.6.1
Fixed
- Resolved issue that prevented cluster-autoscaler from autoscaling workers.
Added
- Added ec2:DescribeNetworkInterfaces to alb-ingress policy
0.6.0
Breaking
- Terraform 0.12 is now required
- Lists are used in place of comma seperated strings.
Added
- Cloudwatch Container Insights support
enable_container_insights
Changed
- Worker group instances are now completely dynamic, you can specify as many instance types as you want with
worker_group_instance_types
- KIAM worker group instances are now completely dynamic, you can specify as many instance types as you want
kiam_instance_types
kiam_vpc_subnets
is now a listkiam_asg_suspended_processes
is now a listkiam_asg_enabled_metrics
is now a listworkers.vpc_subnets
is now a listworkers.suspended_processes
is now a listworkers.enabled_metrics
is now a list
0.5.1
[0.5.1] - 2019-05-21
Fixed
- Resolved issue that resulted in plan failures when
enable_kiam = false
0.5.0
[0.5.0] - 2019-04-15
Breaking
- Clusters will now default to using KIAM for IAM roles. This requires KIAM to be active in your cluster for pods to use IAM role permissions. If you are deploying demo clusters this can be disabled via
enable_kiam
Added
- Added support for EKS Control Plane logs to Cloudwatch. Defaults to all logs enabled.
Fixed
- Resolved issue where Launch Template based ASG's could not be scaled from 0.
0.4.0
[0.4.0] - 2019-03-26
Breaking
- Removed instance_type_3 support. This was hard to use in practice as cluster-autoscaler expects all instances in an ASG to have the same CPU/Memory characteristics. instance_type_2 has been left as for most instance types Amazon has the previous instance generation available which has the same CPU/Memory characteristics. For example M4/M5 Large instances both have 2 vCPU and 8GB of memory available so can share an ASG without impacting cluster-autoscaler.
Fixed
- Add cluster owned tag on the default worker security group so the ELB controller is able to find it when multiple security groups are applied to nodes
Added
- Add support for EKS Public/Private endpoint configuration
- Add support for specifying subnets per ASG to work around PV scheduling limitations
0.3.0: Add some additional worker options (#7)
[0.3.0] - 2019-03-15
Added
- suspended_processes on worker groups
- enabled_metrics on worker groups
0.2.0
[0.2.0] - 2019-03-01
- Added Velero Support
- Resolved issue with using aws_subnet_ids datasource for private_subnets
0.1.0: First release (#1)
[0.1.0] - 2019-02-09
Added Everything