5.0.0 (2025-11-10)
⚠ BREAKING CHANGES
- Upgrade Terraform and AWS min required providers to 1.10 and
6.19respectively (#426)
Features
Note
See the docs/UPGRADE-5.0.md guide for further details
List of backwards incompatible changes
- Minimum supported version of Terraform AWS provider updated to
v6.19to support the latest resources utilized - Minimum supported version of Terraform updated to
v1.10(min supported version for ACM module used within this module) - The underlying
aws_security_group_rulehave been replaced withaws_vpc_security_group_ingress_ruleandaws_vpc_security_group_egress_ruleto allow for more flexibility in defining security group rules. - The attributes used to construct the container definition(s) have been changed from HCL's norm of
snake_casetocamelCaseto match the AWS API. There currently isn't a resource nor data source for the container definition, so one is constructed entirely from HCL in thecontainer-definitionsub-module. This definition is then rendered as JSON when presented to the task definition (or task set) APIs. Previously, the variable names used weresnake_caseand then internally converted tocamelCase. However, this does not allow for using thecontainer-definitionsub-module on its own due to the mismatch between casing. Its probably going to trip a few folks up, but hopefully we'll remove this for a data source in the future. service.task_exec_ssm_param_arnsdefault of["arn:aws:ssm:*:*:parameter/*"]has been removed to prevent unintended permission grants. If you were relying on this default, you will need to explicitly set this variable in your configuration.service.task_exec_secret_arnsdefault of["arn:aws:secretsmanager:*:*:secret:*"]has been removed to prevent unintended permission grants. If you were relying on this default, you will need to explicitly set this variable in your configuration.
Additional changes
Added
- Support for
regionargument to specify the AWS region for the resources created if different from the provider region.
Modified
- The ALB module used within this module has been updated to
v10.2.0 - The ECS cluster and service modules used within this module have been updated to
v6.7.0 - The ACM module used within this module has been updated to
v6.1.1 - The EFS module used within this module has been updated to
v2.0.0 - Variable definitions now contain detailed object types in place of the previously used
anytype
Removed
- None
Variable and output changes
-
Removed variables:
atlantis_gid-> is nowatlantis.group_idwithin theatlantisobject variableatlantis_uid-> is nowatlantis.user_idwithin theatlantisobject variablealb_https_default_action-> replaced byalb.https_default_actionwithin thealbobject variablealb_subnets-> replaced byalb.subnet_idswithin thealbobject variableservice_subnets-> replaced byservice.subnet_idswithin theserviceobject variable- From the
albobject variable:customer_owned_ipv4_pooldesync_mitigation_modedns_record_client_routing_policyenable_tls_version_and_cipher_suite_headersenable_xff_client_portload_balancer_typexff_header_processing_mode
- From the
serviceobject variable:ignore_task_definition_changesalarmsdeployment_controllerdeployment_maximum_percent- Atlantis only supports 1 running instancedeployment_minimum_healthy_percent- Atlantis only supports 1 running instancedesired_count- Atlantis only supports 1 running instanceenable_execute_commandordered_placement_strategyplacement_constraintsscheduling_strategyservice_connect_configurationservice_registriescontainer_definition_defaultsinference_acceleratoripc_modepid_modetask_definition_placement_constraintsproxy_configurationskip_destroyexternal_idscaleforce_deletewait_until_stablewait_until_stable_timeoutenable_autoscalingautoscaling_min_capacityautoscaling_max_capacityautoscaling_policiesautoscaling_scheduled_actions
- From the
atlantisobject variable:essential- now always trueextra_hostsinteractivelinkspseudo_terminalsystem_controls
- From the
efsobject variable:create_backup_policyenable_backup_policycreate_replication_configurationreplication_configuration_destination
-
Renamed variables:
cluster.settings->cluster.setting(singular)cluster.fargate_capacity_providers-> replaced bycluster.default_capacity_provider_strategy