fix: Get on_demand_allocation_strategy from local.workers_group_defaults when deciding to use mixed_instances_policy

[trnubo]

Get on_demand_allocation_strategy from local.workers_group_defaults when deciding to use mixed_instances_policy

PR o'clock

Description

This fix uses the default from local.workers_group_defaults instead of null (the workers_group_defaults_defaults is null anyway) when the dynamic blocks mixed_instances_policy and launch_template are being decided on.

We use many similar worker groups (worker group per AZ, on-demand and spot groups) via worker_groups_launch_template and in an attempt to reduce duplicated config between worker groups we use workers_group_defaults. However setting on_demand_allocation_strategy and override_instance_types in the defaults doesn't have any effect when on_demand_allocation_strategy isn't set in each worker group.

This fix allows on_demand_allocation_strategy to be set in workers_group_defaults which will enable the mixed_instances_policy block and use the default override_instance_types even when neither are defined in the worker group. I don't see the default on_demand_allocation_strategy used anywhere else.

Checklist

• CI tests are passing
• README.md has been updated after any changes to variables and outputs. See https://github.com/terraform-aws-modules/terraform-aws-eks/#doc-generation

[dpiddockcmp]

Should the on_demand_allocation_strategy block around line 100 do a lookup for local.workers_group_defaults["on_demand_allocation_strategy"]? Although at the moment the only valid value is prioritized

[barryib]

Nice catch. Thanks @trnubo

[github-actions]

I'm going to lock this pull request because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues. If you have found a problem that seems related to this change, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.