mirror of
https://github.com/ysoftdevs/terraform-aws-eks.git
synced 2026-03-24 18:31:09 +01:00
If you are trying to recover a cluster that was deleted, the current code will not re-apply the ConfigMap because it is already rendered so kubectl command won't get triggered. This change adds the cluster endpoint (which should be different when spinning up a new cluster even with the same name) so we will force a re-render and cause the kubectl command to run.
10 KiB
10 KiB
Change Log
All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog and this project adheres to Semantic Versioning.
[v2.0.1] - 2019-01-??]
Added
- Write your awesome addition here (by @you)
Changed
- Updated the
update_config_map_aws_authresource to trigger when the EKS cluster endpoint changes. This likely means that a new cluster was spun up so our ConfigMap won't exist (fixes #234) (by @elatt) - Removed invalid action from worker_autoscaling iam policy (by @marcelloromani)
- Fixed zsh-specific syntax in retry loop for aws auth config map (by @marcelloromani)
- Fix: fail deployment if applying the aws auth config map still fails after 10 attempts (by @marcelloromani)
[v2.0.0] - 2018-12-14]
Added
- (Breaking Change) New input variables
map_accounts_count,map_roles_countandmap_users_countto allow using computed values as part ofmap_accounts,map_rolesandmap_usersconfigs (by @chili-man on behalf of OpenGov). - (Breaking Change) New variables
cluster_create_security_groupandworker_create_security_groupto stopvalue of 'count' cannot be computederror. - Added ability to choose local-exec interpreter (by @rothandrew)
Changed
- Added
--with-aggregate-type-defaultsoption to terraform-docs (by @max-rocket-internet) - Updated AMI ID filtering to only filter AMIs from current cluster k8s version (by @max-rocket-internet)
- Added
pre-commit-terraformgit hook to automatically create documentation of inputs/outputs (by @antonbabenko) - Travis fixes (by @RothAndrew)
- Fixed some Windows compatibility issues (by @RothAndrew)
[v1.8.0] - 2018-12-04]
Added
- Support for using AWS Launch Templates to define autoscaling groups (by @skang0601)
suspended_processestoworker_groupsinput (by @bkmeneguello)target_group_arnstoworker_groupsinput (by @zihaoyu)force_detach_policiestoaws_iam_roleclusterandworkers(by @marky-mark)- Added sleep while trying to apply the kubernetes configurations if failed, up to 50 seconds (by @rmakram-ims)
cluster_create_security_groupandworker_create_security_group. This allows using computed cluster and worker security groups. (by @rmakram-ims)
Changed
- new variables worker_groups_launch_template and worker_group_count_launch_template (by @skang0601)
- Remove aws_iam_service_linked_role (by @max-rocket-internet)
- Adjust the order and correct/update the ec2 instance type info. (@chenrui333)
- Removed providers from
main.tf. (by @max-rocket-internet) - Removed
configure_kubectl_sessionreferences in documentation #171 (by @dominik-k)
[v1.7.0] - 2018-10-09]
Added
- Worker groups can be created with a specified IAM profile. (from @laverya)
- exposed
aws_eks_clustercreate and destroy timeouts (by @RGPosadas) - exposed
placement_tenancyfor autoscaling group (by @monsterxx03) - Allow port 443 from EKS service to nodes to run
metrics-server. (by @max-rocket-internet)
Changed
- fix default worker subnets not working (by @erks)
- fix default worker autoscaling_enabled not working (by @erks)
- Cosmetic syntax changes to improve readability. (by @max-rocket-internet)
- add
protect_from_scale_into solve issue #134 (by @kinghajj)
[v1.6.0] - 2018-09-04]
Added
- add support for
amazon-eks-node-*AMI with bootstrap script (by @erks) - expose
kubelet_extra_argsworker group option (replacingkubelet_node_labels) to allow specifying arbitrary kubelet options (e.g. taints and labels) (by @erks) - add optional input
worker_additional_security_group_idsto allow one or more additional security groups to be added to all worker launch configurations - #47 (by @hhobbsh @mr-joshua) - add optional input
additional_security_group_idsto allow one or more additional security groups to be added to a specific worker launch configuration - #47 (by @mr-joshua)
Changed
- allow a custom AMI to be specified as a default (by @erks)
- bugfix for above change (by @max-rocket-internet)
- Breaking change Removed support for
eks-worker-*AMI. The cluster specifying a custom AMI based off ofeks-worker-*AMI will have to rebuild the AMI fromamazon-eks-node-*. (by @erks) - Breaking change Removed
kubelet_node_labelsworker group option in favor ofkubelet_extra_args. (by @erks)
[v1.5.0] - 2018-08-30]
Added
- add spot_price option to aws_launch_configuration
- add enable_monitoring option to aws_launch_configuration
- add t3 instance class settings
- add aws_iam_service_linked_role for elasticloadbalancing. (by @max-rocket-internet)
- Added autoscaling policies into module that are optionally attached when enabled for a worker group. (by @max-rocket-internet)
Changed
- Breaking change Removed
workstation_cidrvariable, http callout and unnecessary security rule. (by @dpiddockcmp) If you are upgrading from 1.4 you should fix state after upgrade:terraform state rm module.eks.data.http.workstation_external_ip - Can now selectively override keys in
workers_group_defaultsvariable rather than callers maintaining a duplicate of the whole map. (by @dpiddockcmp)
[v1.4.0] - 2018-08-02]
Added
- manage eks workers' root volume size and type.
workers_asg_namesadded to outputs. (kudos to @laverya)- New top level variable
worker_group_countadded to replace the use oflength(var.worker_groups). This allows using computed values as part of worker group configs. (complaints to @laverya)
[v1.3.0] - 2018-07-11]
Added
- New variables
map_accounts,map_rolesandmap_usersin order to manage additional entries in theaws-authconfigmap. (by @max-rocket-internet) - kubelet_node_labels worker group option allows setting --node-labels= in kubelet. (Hat-tip, @bshelton229 👒)
worker_iam_role_arnadded to outputs. Sweet, @hatemosphere 🔥
Changed
- Worker subnets able to be specified as a dedicated list per autoscaling group. (up top, @bshelton229 🙏)
[v1.2.0] - 2018-07-01]
Added
- new variable
pre_userdataadded to worker launch configuration allows to run scripts before the plugin does anything. (W00t, @jimbeck 🦉)
Changed
- kubeconfig made much more flexible. (Bang up job, @sdavids13 💥)
- ASG desired capacity is now ignored as ASG size is more effectively handed by k8s. (Thanks, @ozbillwang 💇♂️)
- Providing security groups didn't behave as expected. This has been fixed. (Good catch, @jimbeck 🔧)
- workstation cidr to be allowed by created security group is now more flexible. (A welcome addition, @jimbeck 🔐)
[v1.1.0] - 2018-06-25]
Added
- new variable
worker_sg_ingress_from_portallows to change the minimum port number from which pods will accept communication (Thanks, @ilyasotkov 👏). - expanded on worker example to show how multiple worker autoscaling groups can be created.
- IPv4 is used explicitly to resolve testing from IPv6 networks (thanks, @tsub 🙏).
- Configurable public IP attachment and ssh keys for worker groups. Defaults defined in
worker_group_defaults. Nice, @hatemosphere 🌂 worker_iam_role_namenow an output. Sweet, @artursmet 🕶️
Changed
- IAM test role repaired by @lcharkiewicz 💅
kube-proxyrestart no longer needed in userdata. Good catch, @hatemosphere 🔥- worker ASG reattachment wasn't possible when using
name. Moved toname_prefixto allow recreation of resources. Kudos again, @hatemosphere 🐧
[v1.0.0] - 2018-06-11]
Added
- security group id can be provided for either/both of the cluster and the workers. If not provided, security groups will be created with sufficient rules to allow cluster-worker communication. - kudos to @tanmng on the idea ⭐
- outputs of security group ids and worker ASG arns added for working with these resources outside the module.
Changed
- Worker build out refactored to allow multiple autoscaling groups each having differing specs. If none are given, a single ASG is created with a set of sane defaults - big thanks to @kppullin 🥨
[v0.2.0] - 2018-06-08]
Added
- ability to specify extra userdata code to execute following kubelet services start.
- EBS optimization used whenever possible for the given instance type.
- When
configure_kubectl_sessionis set to true the current shell will be configured to talk to the kubernetes cluster using config files output from the module.
Changed
- files rendered from dedicated templates to separate out raw code and config from
hcl workers_ami_idis now made optional. If not specified, the module will source the latest AWS supported EKS AMI instead.
[v0.1.1] - 2018-06-07]
Changed
- Pre-commit hooks fixed and working.
- Made progress on CI, advancing the build to the final
kitchen teststage before failing.
[v0.1.0] - 2018-06-07
Added
- Everything! Initial release of the module.
- added a local variable to do a lookup against for a dynamic value in userdata which was previously static. Kudos to @tanmng for finding and fixing bug #1!