edge.26-04-05

Add alias to aws_organization

The aws_organization module now manages the IAM account alias for the management AWS account via a new required alias variable.

Add an alias input to your aws_organization module configuration before the next terragrunt apply:

inputs = {
  alias = "my-org-management" # A human-readable string for the management account
  # ... other existing inputs
}

Consolidate Contact Information Variables

The contact information variables on aws_account and aws_registered_domains have been consolidated to single objects for each contact.

The inputs to aws_dns_zones have been consolidated to a single domains object for better per-domain configuration.

Replace the individual contact field variables (e.g., contact_first_name, contact_last_name, etc.) with the new consolidated contact objects in your aws_account and aws_registered_domains module configurations. Refer to the updated module reference docs for the exact object structure.

Migrate aws_dns_zones Inputs to domains Object

The inputs to aws_dns_zones have been consolidated from separate domain list variables into a single domains object that allows per-domain granular configuration.

Migrate your aws_dns_zones inputs so that each key in the domains object is a domain name and the value contains per-domain configuration. See the module reference for the new input schema.

Rename Dedicated Cluster DNS Zone

Every cluster now has a dedicated DNS zone for hosting control-plane utilities. kube_domain is now a required configuration value and should be set in the region.yaml for every region that houses a Kubernetes cluster. The value must be a subdomain of a domain available to the environment.

1. Add a kube_domain field to region.yaml for every region that contains a Kubernetes cluster. The value must be a subdomain of a domain already available in the environment (e.g., kube.example.com if example.com is managed).

2. Deploy the corresponding DNS zone via aws_dns_zones before applying cluster modules.

Remove Secondary Provider from tf_bootstrap_resources

The aws.secondary provider alias and cross-region DynamoDB replica have been removed from tf_bootstrap_resources.

1. Remove the aws.secondary provider alias from your tf_bootstrap_resources Terragrunt configuration if present.
2. Re-apply tf_bootstrap_resources to remove the cross-region DynamoDB replica from your state lock table.

Rename Backup Vault in tf_bootstrap_resources

The backup vault name in tf_bootstrap_resources now has a unique suffix to prevent conflicts. Before applying the updated module, manually delete the existing backup vault named terraform-<env_name>. Delete all recovery points first, then the vault.

Migrate Authentik Token to region.secrets.yaml

The pf sso add command previously stored the Authentik API token as authentikUserToken inside authentik_core_resources/secrets.yaml. The token has been relocated to region.secrets.yaml under the standardized key authentik_token.

If you have already run pf sso add and have an existing token stored in <cluster-path>/authentik_core_resources/secrets.yaml under the key authentikUserToken, migrate it to the region-level file:

sops --set '["authentik_token"] "your-token-here"' <region-path>/region.secrets.yaml

Ensure authentik_token is set in region.secrets.yaml (SOPS-encrypted) before re-applying any Authentik modules.

Update Authentik URL to sso.<domain>

The Authentik subdomain has been standardized from authentik.<domain> to sso.<domain>.

If you previously deployed Authentik at authentik.<domain>, you must:

1. Update the authentik_url value in global.yaml from authentik.<domain> to sso.<domain>.
2. Create a DNS record for sso.<domain> pointing to the Authentik ingress.
3. Re-apply kube_authentik so the module picks up the new domain.
4. After verifying the new domain works, remove the old authentik.<domain> DNS record.

Configure Authentik Organization Name

We now create the Authentik email template in kube_authentik. The organization name is now a required input. Use the new organization_name output from the kube_authentik module as an input to the authentik_core_resources module.

1. Add the organization_name variable to your kube_authentik module configuration.

2. Wire the organization_name output from kube_authentik into the authentik_core_resources module as an input via a dependency block:

   dependency "authentik" {
     config_path = "../kube_authentik"
   }
   
   inputs = {
     organization_name = dependency.authentik.outputs.organization_name
     # ... other inputs
   }

Install KEDA

We now include KEDA in our base Panfactum clusters and our modules assume that you have it installed. See the instructions here.

Remove Node Image Cache Modules

The kube_node_image_cache and kube_node_image_cache_controller modules have been removed entirely.

1. Destroy any active kube_node_image_cache module deployments.
2. Destroy any active kube_node_image_cache_controller module deployments.
3. Remove the module directories for kube_node_image_cache and kube_node_image_cache_controller from your Terragrunt configuration.
4. Remove the following input variables from any module configurations that set them:
   • node_image_cached_enabled — from kube_airbyte, kube_alloy, kube_argo_event_bus, kube_authentik, kube_aws_ebs_csi, kube_cloudnative_pg, kube_gha_runners, kube_ingress_nginx, kube_linkerd, kube_monitoring, kube_nats, kube_opensearch, kube_pg_cluster, kube_redis_sentinel, kube_vault
   • node_image_cache_enabled — from any module that sets it
   • image_prepull_enabled and image_pin_enabled — from container spec blocks in kube_pod, kube_deployment, kube_daemon_set, kube_stateful_set, kube_cron_job, and kube_job
   • panfactum_node_image_cache_enabled — from kube_policies

Upgrade Kubernetes to 1.33

The default Kubernetes version has been upgraded to 1.33 (from 1.30, incrementing through 1.31 and 1.32 in this release).

1. Review the Kubernetes 1.33 changelog for any deprecated APIs or behavior changes that affect your workloads.
2. If you pin kube_version explicitly in aws_eks, update it to 1.33. Otherwise, re-apply aws_eks to trigger the upgrade.

No additional action is required for the bundled compatibility fixes (EBS CSI driver pinning, Descheduler config migration, and Ingress-nginx annotation risk level).

Migrate kube_cert_manager and kube_cert_issuers to kube_certificates

kube_cert_manager and kube_cert_issuers have been consolidated into a single kube_certificates module.

1. Create a kube_certificates directory as a sibling to the kube_cert_manager and kube_cert_issuers directories.

2. Create a terragrunt.hcl file in the kube_certificates directory with the following contents:

   include "panfactum" {
     path   = find_in_parent_folders("panfactum.hcl")
     expose = true
   }
   
   terraform {
     source = include.panfactum.locals.pf_stack_source
   }
   
   dependency "vault_core" {
     config_path  = "../vault_core_resources"
     skip_outputs = true
   }
   
   inputs = {
     alert_email = "..." # Copy from kube_cert_issuers
   }

3. From the region directory, run the following:

   (cd kube_cert_issuers; terragrunt state pull > state.json);
   (cd kube_cert_manager; terragrunt state pull > state.json);
   jq -s '
     .[0] as $f1
     | .[1] as $f2
     | $f1
       | .outputs = ($f1.outputs + $f2.outputs)
       | .resources = ($f1.resources + $f2.resources)
   ' kube_cert_issuers/state.json kube_cert_manager/state.json > kube_certificates/state.json;
   jq 'del(
     .resources[]
     | select(
       .type == "pf_kube_labels"
       and (has("module") | not)
     )
   )' kube_certificates/state.json > tmp && mv tmp kube_certificates/state.json;
   rm kube_cert_issuers/state.json;
   rm kube_cert_manager/state.json;

4. Update the version of the Panfactum framework you are using.

5. Navigate to the kube_certificates directory.

   1. Run terragrunt init.
   2. Run terragrunt state push state.json && rm state.json.
   3. Run terragrunt apply and review the changes. There should be only a few resources that will be replaced.

6. Remove the kube_cert_issuers and kube_cert_manager directories.

7. Replace any references to kube_cert_issuers and kube_cert_manager with kube_certificates in your code.

Review pull_through_cache_enabled Default

pull_through_cache_enabled now defaults to true for kube_nats, kube_pg_cluster, kube_redis_sentinel, kube_stateful_set, and kube_deployment. If ECR pull-through caching is not configured in your environment, image pulls for these modules will fail.

Choose one of the following options:

• Option A: Deploy the aws_ecr_pull_through_cache module to enable pull-through caching before re-applying these modules.
• Option B: Explicitly set pull_through_cache_enabled = false on each of kube_nats, kube_pg_cluster, kube_redis_sentinel, kube_stateful_set, and kube_deployment in your module configurations.

Remove update_type from kube_cron_job

The update_type variable has been removed from kube_cron_job.

Remove any update_type input from your kube_cron_job module configurations to avoid a Terraform error.

Review burstable_nodes_enabled Default

burstable_nodes_enabled now defaults to true. If you explicitly set it to false, no action is needed. Otherwise, confirm your workloads are compatible with burstable (T-family) instances.

Update PF_SKIP_REPO_CHECK Environment Variable

The PF_SKIP_REPO_CHECK environment variable has been renamed to PF_SKIP_CHECK_REPO_SETUP.

If you previously set PF_SKIP_REPO_CHECK=1 to skip the repo setup check, update your environment, CI pipelines, and scripts to use PF_SKIP_CHECK_REPO_SETUP=1 instead.

Update pf install-cluster References

The pf install-cluster CLI command has been renamed to pf cluster add.

Replace all invocations of pf install-cluster with pf cluster add in your scripts, CI pipelines, and runbooks.

Initialize the pf CLI

This release adds the new pf CLI tool. To begin using it:

1. Complete all migration steps for the breaking changes above.
2. Run pf devshell sync. Ensure this completes successfully before proceeding.
3. Run terragrunt apply on all modules (or terragrunt run-all apply).