diff --git a/README.md b/README.md index 54c4ff1..9fef742 100644 --- a/README.md +++ b/README.md @@ -25,12 +25,12 @@ - [5.2 Automatic TEST and STAGE Deployment](#52-automatic-test-and-stage-deployment) - [5.3 Manual Deployment through PROD](#53-manual-deployment-through-prod) - [5.4 Local Development, Validation, and Manual Deployment](#54-local-development-validation-and-manual-deployment) - - [5.5 TEST and STAGE Destruction and Recreation](#55-test-and-stage-destruction-and-recreation) + - [5.5 Combined TEST and STAGE Destruction and Recreation](#55-combined-test-and-stage-destruction-and-recreation) - [5.6 PROD Destruction and Recreation](#56-prod-destruction-and-recreation) - [5.7 Individual TEST and STAGE Destruction and Recreation](#57-individual-test-and-stage-destruction-and-recreation) 6. [Destroying Resources and Cleanup](#6-destroying-resources-and-cleanup) - - [6.1 Automated Destruction of TEST and STAGE](#61-automated-destruction-of-test-and-stage) - - [6.2 Automated Destruction of PROD](#62-automated-destruction-of-prod) + - [6.1 Manually Triggered Destruction of TEST and STAGE](#61-manually-triggered-destruction-of-test-and-stage) + - [6.2 Manually Triggered Destruction of PROD](#62-manually-triggered-destruction-of-prod) - [6.3 Independent TEST and STAGE Destruction](#63-independent-test-and-stage-destruction) - [6.4 Manually Destroy an Individual Frontend Stack](#64-manually-destroy-an-individual-frontend-stack) - [6.5 Manually Destroy All Stacks in One Environment](#65-manually-destroy-all-stacks-in-one-environment) @@ -73,44 +73,49 @@ The backend and database *subnets* are created as part of the network topology; | Centralized CI tool versions | `.opentofu-version`, `.tflint-version`, `requirements-checkov.txt` | | Independently tagged releases of reusable modules | **Not implemented**; modules are versioned with repository commits | -The destruction and recreation workflows were developed after the original November 1 submission. This README describes the current repository state rather than claiming that post-submission changes were part of the original ZIP. Runner availability and instructor permissions still require verification through GitHub and OpenStack administration. +The destruction and recreation workflows were developed after the original November 1 submission. This README describes the current repository state rather than claiming that post-submission changes were part of the original ZIP. The self-hosted runner has been verified as an active, persistent +`systemd` service and is available in GitHub Actions. + ## 2. Architecture and repository structure ```text assigment2-iac/ -├── .github/workflows/ -│ ├── ci.yml # Formatting, validation, TFLint, Checkov -│ ├── deploy.yml # TEST → STAGE → conditional PROD -│ ├── destroy.yml # Manually triggered TEST + STAGE destruction -│ ├── destroy-test.yml # Manually triggered TEST-only destruction -│ ├── destroy-stage.yml # Manually triggered STAGE-only destruction -│ ├── destroy-prod.yml # Manually triggered, approval-gated PROD destruction -│ └── runner-test.yml # Manually triggered runner health check +├── .github/ +│ └── workflows/ +│ ├── ci.yml # Formatting, validation, TFLint, Checkov +│ ├── deploy.yml # Automatic TEST/STAGE, manual PROD +│ ├── destroy.yml # Combined TEST and STAGE destruction +│ ├── destroy-test.yml # TEST-only destruction +│ ├── destroy-stage.yml # STAGE-only destruction +│ └── destroy-prod.yml # Approval-protected PROD destruction ├── bootstrap/ -│ ├── cloud_init/runner.yaml # Runner VM operating-system packages -│ └── *.tf # Runner network, router, VM, security groups +│ ├── cloud_init/ +│ │ └── runner.yaml # Runner VM initialization +│ └── *.tf # Runner infrastructure configuration ├── environments/ │ ├── test/ -│ │ ├── network/ # TEST network state and configuration -│ │ └── frontend/ # TEST frontend state and configuration +│ │ ├── network/ # TEST network stack +│ │ └── frontend/ # TEST frontend stack │ ├── stage/ -│ │ ├── network/ -│ │ └── frontend/ +│ │ ├── network/ # STAGE network stack +│ │ └── frontend/ # STAGE frontend stack │ └── prod/ -│ ├── network/ -│ └── frontend/ +│ ├── network/ # PROD network stack +│ └── frontend/ # PROD frontend stack ├── modules/ -│ ├── network/ # Reusable OpenStack network resources -│ └── vm/ # Reusable VM, SG, IP, optional volume attach -├── .opentofu-version # Pinned OpenTofu version (1.12.6) -├── .tflint-version # Pinned TFLint version -├── requirements-checkov.txt # Pinned Checkov dependency -├── .tflint.hcl +│ ├── network/ # Reusable OpenStack network module +│ └── vm/ # Reusable VM and security group module +├── screenshots/ +├── .opentofu-version # Pinned OpenTofu version (1.12.6) +├── .tflint-version # Pinned TFLint version +├── requirements-checkov.txt # Pinned Checkov dependency +├── .tflint.hcl # TFLint configuration ├── .gitignore └── README.md ``` + **Independent destructive-operation entry points:** `destroy-test.yml` and `destroy-stage.yml` provide selective cleanup, while the verified `destroy.yml` remains available for combined non-production cleanup. `destroy-prod.yml` is independent and approval-protected. Every environment has its own network and frontend root module, its own `deployment.tfvars`, and its own `.terraform.lock.hcl`. The frontend stack uses `terraform_remote_state` to read the network ID and frontend subnet ID from the corresponding environment's network state. Therefore **network must be created before frontend, and frontend must be destroyed before network**. @@ -135,7 +140,7 @@ The `bootstrap/` root module creates the dedicated self-hosted runner VM and its 2. Make changes to the appropriate environment, module, or workflow files. 3. Run local formatting, validation, and relevant lint/security checks (sections 3.2 and 5.4). 4. Commit and push the feature branch. -5. Open a pull request targeting `main`. Review the OpenTofu changes and verify the CI job passes before merging. +5. Open a pull request targeting `main` and optionally request a review from a repository collaborator. Review the OpenTofu changes and verify the CI job passes before merging. 6. Merge the reviewed change into `main`. The successful push CI run triggers the deployment workflow. #### Protect main branch ruleset @@ -172,7 +177,9 @@ CI is triggered on **pull requests targeting `main`**, **pushes to `main`**, or The tool-version files are tracked in Git. They make deliberate upgrades easy to review, but do **not** automatically update themselves. A failed check causes CI to fail. Only a successful **push-originated** CI run on `main` meets the automatic deployment workflow's trigger condition; a pull-request CI run does not deploy infrastructure. -**Security scan findings and resolution:** Enabling Checkov's `secrets` framework initially produced six `CKV_SECRET_6` (high-entropy string) findings on `secret_suffix` values in the Kubernetes backend `providers.tf` files for TEST, STAGE, and PROD. These are deterministic **OpenTofu remote-state identifiers**, not credentials. The six findings were reviewed and addressed with narrowly scoped `checkov:skip=CKV_SECRET_6` comments on the relevant backend values rather than disabling the rule globally. After reformatting the modified `.tf` files with `tofu fmt`, CI succeeded. Secret scanning remains enabled and continues to check other repository content. Checkov provides policy coverage for supported IaC/workflow issues but does not guarantee comprehensive checks of OpenStack-specific security groups and resources. +**GitHub Actions workflow finding (`CKV_GHA_7`):** Checkov identified `CKV_GHA_7` because the deployment workflow initially accepted `workflow_dispatch` inputs. These inputs were removed to satisfy the security check. Instead, the deployment workflow distinguishes between automatic and manually triggered executions: successful CI runs on `main` automatically deploy TEST and STAGE, while a manually triggered deployment proceeds through TEST, STAGE, and approval-protected PROD. This preserves the required deployment behavior without relying on manual workflow inputs. + +**Secret-scanning findings (`CKV_SECRET_6`):** Enabling Checkov's `secrets` framework initially produced six `CKV_SECRET_6` (high-entropy string) findings on `secret_suffix` values in the Kubernetes backend `providers.tf` files for TEST, STAGE, and PROD. These are deterministic **OpenTofu remote-state identifiers**, not credentials. The six findings were reviewed and addressed with narrowly scoped `checkov:skip=CKV_SECRET_6` comments on the relevant backend values rather than disabling the rule globally. After reformatting the modified `.tf` files with `tofu fmt`, CI succeeded. Secret scanning remains enabled and continues to check other repository content. Checkov provides policy coverage for supported IaC/workflow issues but does not guarantee comprehensive checks of OpenStack-specific security groups and resources. ### 3.3 Deployment — `.github/workflows/deploy.yml` @@ -198,20 +205,24 @@ Manual "Deploy Infrastructure" run from main `deploy.yml` is triggered by a completed CI workflow on `main` or by `workflow_dispatch`. The TEST job checks that CI succeeded and was triggered by a push to `main`, or that the deployment was manually dispatched from `main`. STAGE has `needs: deploy-test`, so it runs only after TEST succeeds. PROD has `needs: deploy-stage` and an additional condition requiring a **manual `workflow_dispatch` from `main`**. There is no `deploy_prod` checkbox in this version: **every manual run of Deploy Infrastructure from `main` attempts PROD after TEST and STAGE**. -For each environment, the deployment job checks OpenStack credentials, installs checksum-verified `kubectl` and `kOps`, obtains a short-lived **1-hour Kubernetes admin kubeconfig**, verifies access to the `iac` namespace, initializes the backends, then processes **network before frontend**. OpenTofu saves each plan and inspects the plan JSON with `jq`. The pipeline only allows `no-op`, `read`, `create`, and `update` resource actions; it rejects planned deletions and replacements. It then applies the exact saved plan, or skips application when there are no changes. Temporary plans, logs, and kubeconfig files are removed in cleanup steps. +For each environment, the deployment job checks OpenStack credentials, installs checksum-verified `kubectl` and `kOps`, obtains a short-lived **1-hour Kubernetes admin kubeconfig**, verifies access to the `iac` namespace, initializes the backends, then processes **network before frontend**. OpenTofu saves each plan and inspects the plan JSON with `jq`. The pipeline only allows `no-op`, `read`, `create`, and `update` resource actions; it rejects planned deletions and replacements. + +**Design decision:** Destructive changes are blocked during normal deployments to prevent accidental resource deletion or replacement, which could cause service interruptions or data loss. Infrastructure destruction is instead handled through dedicated, manually triggered workflows, ensuring that destructive operations require explicit operator action. + +The pipeline then applies the exact saved plan, or skips application when there are no changes. Temporary plans, logs, and kubeconfig files are removed in cleanup steps. The TEST `Initialize TEST remote state` step no longer requires pre-existing managed resources. It initializes the network and frontend Kubernetes backends even when the state is empty, allowing TEST to be provisioned initially or recreated following a complete destroy. **All three environments have now successfully completed deploy → destroy → recreate tests through GitHub Actions.** PROD was recreated through a fresh manual deployment run after the dedicated PROD destruction succeeded. ### 3.4 Combined TEST and STAGE destruction — `.github/workflows/destroy.yml` -The separate **Destroy Infrastructure (TEST, STAGE)** workflow follows the Lab 7 pattern and is **manually triggered** with `workflow_dispatch` from `main`. A GitHub Actions **matrix** creates one job for each of these environment mappings: +The separate **Destroy Infrastructure (TEST, STAGE)** workflow is based on the GitHub Actions matrix pattern demonstrated in [Lab 7](https://git.ntnu.no/jornskj-org/iikg3005_h26_lab7). It is **manually triggered** using `workflow_dispatch` from `main`. A GitHub Actions **matrix strategy** creates a separate job for each of the following environment configurations: | GitHub environment | Configuration path | State suffixes | | --- | --- | --- | | `TEST` | `environments/test/` | `assignment2-test-frontend`, `assignment2-test-network` | | `STAGE` | `environments/stage/` | `assignment2-stage-frontend`, `assignment2-stage-network` | -The matrix currently **does not include PROD**. Each job selects the corresponding GitHub environment for its credentials, verifies its configuration and backend state suffixes, and operates only on that environment's two stacks. +The matrix **does not include PROD**. Each job selects the corresponding GitHub environment for its credentials, verifies its configuration and backend state suffixes, and operates only on that environment's two stacks. The workflow runs on the existing self-hosted runner. It installs OpenTofu from `.opentofu-version`, installs checksum-verified `kubectl` and kOps, exports a temporary one-hour Kubernetes admin kubeconfig, and checks access to the `iac` namespace. For **each environment**, it performs: @@ -228,9 +239,9 @@ The matrix uses `max-parallel: 1`, so only one environment job executes at a tim ### 3.5 Dedicated PROD destruction — `.github/workflows/destroy-prod.yml` -The **Destroy Infrastructure (PROD)** workflow is a separate, manually dispatched entry point available from `main`. It does **not** use the TEST/STAGE matrix. Its job uses `environment: PROD`, so the **existing required human approval gate** must be satisfied before any destruction steps execute. Authorized reviewers are the repository owner, Jørn, and Pau; self-approval depends on the configured environment rules. +The **Destroy Infrastructure (PROD)** workflow is a separate, manually dispatched entry point available from `main`. It does **not** use the TEST/STAGE matrix. Its job uses `environment: PROD`, so the **existing required human approval gate** must be satisfied before any destruction steps execute. Authorized reviewers are the repository owner, Jørn, and Pau. -Once approved, the job uses PROD-scoped OpenStack credentials, validates Kubernetes remote-state suffixes `assignment2-prod-frontend` and `assignment2-prod-network`, initializes the backends, and applies saved `tofu plan -destroy` plans **frontend first, then network**. It shares `infrastructure-deployment` concurrency with deployment and other destruction workflows. It does not target TEST, STAGE, the self-hosted runner, or the Kubernetes state cluster. +Once approved, the job uses PROD-scoped OpenStack credentials, validates Kubernetes remote-state suffixes `assignment2-prod-frontend` and `assignment2-prod-network`, initializes the backends, and applies saved `tofu plan -destroy` plans, **frontend first, then network**. It shares `infrastructure-deployment` concurrency with deployment and other destruction workflows. It does not target TEST, STAGE, the self-hosted runner, or the Kubernetes state cluster. **Verified:** The PROD destruction run succeeded after environment approval. A subsequent **new, manually dispatched** `Deploy Infrastructure` run successfully recreated PROD through the TEST → STAGE → approved PROD sequence. @@ -245,7 +256,10 @@ Two dedicated, manually triggered workflows enable independent cleanup without c Both workflows require a manual `workflow_dispatch` from `main` and specify their GitHub environment and target path explicitly; there is **no dispatch input or matrix** that could select the other environment. They reuse the verified pattern: environment-scoped credentials, Kubernetes backend/state validation, a saved OpenTofu destroy plan, **frontend → network** ordering, the shared `infrastructure-deployment` concurrency group, and cleanup of temporary files. -**Verified:** TEST-only destruction removed TEST while STAGE and PROD remained intact; TEST was recreated through deployment. STAGE-only destruction then removed STAGE while TEST and PROD remained intact; STAGE was recreated successfully. The combined `destroy.yml` still provides intentional cleanup of both non-production environments. Keeping these workflows separate offers explicit operator scope at the cost of duplicated implementation steps. A future `workflow_call` reusable workflow could centralize the common logic. +**Design decision:** The original `destroy.yml` was retained to support combined cleanup of TEST and STAGE through a single workflow. The dedicated workflows provide more targeted destruction while preserving the existing, tested matrix-based implementation. This offers operational flexibility and explicit environment selection at the cost of duplicated workflow logic. A future reusable `workflow_call` workflow could reduce this duplication. + +**Verified:** TEST-only destruction removed TEST while STAGE and PROD remained intact; TEST was recreated through deployment. STAGE-only destruction then removed STAGE while TEST and PROD remained intact; STAGE was recreated successfully. + ## 4. Prerequisites and credentials The workflow assumes the following pre-existing infrastructure and permissions: @@ -280,11 +294,32 @@ The workflow additionally sets `OS_AUTH_TYPE=v3applicationcredential`, `OS_IDENT ### 5.1 Runner verification -1. Verify the runner VM is powered on in OpenStack. -2. Verify it appears **Online** in repository **Settings → Actions → Runners** with label `assignment2`. -3. Go to **Actions → Runner Test → Run workflow**, select `main`, and confirm that the runner and OpenTofu version checks succeed. +The self-hosted GitHub Actions runner is hosted on the `assignment2-runner` VM in OpenStack. + +1. Verify that the VM is powered on in OpenStack. +2. Check the [GitHub Actions runner status](https://git.ntnu.no/einalien/assigment2-iac/settings/actions/runners) and confirm that `assignment2-runner` is **Online** or **Idle**, with the label `assignment2`. +3. Verify successful CI/CD workflow executions under **Actions → CI** or **Deploy Infrastructure**. + +The runner is configured as a persistent `systemd` service: + +`actions.runner._services.assignment2-runner.service` + +The service has been verified as `enabled` (automatic startup on boot) and `active` (currently running). + +![Self-hosted runner online](screenshots/runner.png) + +The runner infrastructure is provisioned through `bootstrap/`, while runner registration and service configuration are handled separately. The VM must remain powered on and connected to GitHub Actions for continuous availability. -The bootstrap configuration for the runner is in `bootstrap/`. An administrator with OpenStack credentials can initialize it locally using a private `terraform.tfvars` based on `bootstrap/terraform.tfvars.example`, then run `tofu init`, `tofu plan`, and `tofu apply`. Preserve the bootstrap state; do not destroy the runner VM while relying on it for CI/CD. +#### Instructor access + +Jørn (`jornskj`) and Pau (`paut`) have direct collaborator access +to the repository, allowing them to execute GitHub Actions workflows. + +Workflows can be triggered through the repository's **Actions** tab +by selecting the desired workflow, choosing **Run workflow**, and +selecting the `main` branch. + +Direct SSH access to the runner VM is not required to execute workflows. ### 5.2 Automatic TEST and STAGE deployment @@ -298,7 +333,7 @@ The bootstrap configuration for the runner is in `bootstrap/`. An administrator ### 5.3 Manual deployment through PROD -1. Open the repository on `git.ntnu.no` and select **Actions → Deploy Infrastructure**. +1. Open the repository [einalien/assigment2-iac](https://git.ntnu.no/einalien/assigment2-iac) and select **Actions → Deploy Infrastructure**. 2. Click **Run workflow** and select **`main`**. This manually invokes the pipeline even if no new commit has been pushed. 3. Wait for **Deploy TEST** and **Deploy STAGE** to succeed. 4. The configured PROD approval gate pauses the production job. An authorized reviewer selects **Review deployments → PROD → Approve and deploy**. @@ -367,7 +402,9 @@ done OpenTofu prompts for confirmation of each apply. Choose the appropriate credentials **before** running this sequence, especially if the environments use different OpenStack application credentials. Local plans may include destructive changes; unlike `deploy.yml`, this command sequence does not implement the CI pipeline's JSON action filter. Review each plan carefully before approval. -### 5.5 TEST and STAGE destruction and recreation +### 5.5 Combined TEST and STAGE destruction and recreation + +**Warning:** This workflow destroys infrastructure in **both TEST and STAGE**. To destroy only one environment, use the dedicated workflows described in [Section 5.7](#57-individual-test-and-stage-destruction-and-recreation). The complete **deploy → destroy → recreate** lifecycle has been successfully verified for **both TEST and STAGE** through GitHub Actions. @@ -405,11 +442,13 @@ Each destroy job removes its own **frontend first, then network**. Do not use th ## 6. Destroying resources and cleanup +Infrastructure destruction is not triggered automatically by deployments or changes to `main`. All destruction workflows require explicit manual execution through GitHub Actions using `workflow_dispatch`. Once triggered, the workflows automatically initialize the remote state, generate and apply saved OpenTofu destroy plans, and clean up temporary files. PROD destruction additionally requires human approval before execution. + **Current status:** All four destruction workflows have been implemented and successfully executed: `.github/workflows/destroy.yml` (combined TEST + STAGE), `.github/workflows/destroy-test.yml` (TEST only), `.github/workflows/destroy-stage.yml` (STAGE only), and `.github/workflows/destroy-prod.yml` (approval-protected PROD only). All three environments have successfully been recreated through GitHub Actions. Normal `deploy.yml` continues to reject deletions and replacements. The CLI procedures below also support targeted individual-stack cleanup. **Warning:** Destruction deletes real OpenStack resources. Obtain authorization, capture the required deployment evidence, protect all remote state, and confirm the target environment before proceeding. Never destroy `bootstrap/` while it hosts the required runner. Never destroy the external kOps cluster because the six stack states depend on it. -### 6.1 Automated destruction of TEST and STAGE +### 6.1 Manually triggered destruction of TEST and STAGE 1. Open **Actions → Destroy Infrastructure (TEST, STAGE)** in the repository. 2. Choose **Run workflow**, select **`main`**, and confirm you intend to remove infrastructure in **both TEST and STAGE**. @@ -419,7 +458,7 @@ Each destroy job removes its own **frontend first, then network**. Do not use th **Scope and safety:** There is no environment or stack selector in the current GitHub Actions destroy workflow. `max-parallel: 1` serializes the matrix jobs but does not guarantee which environment executes first; `fail-fast: false` means one job's failure does not stop the other from being attempted. These jobs apply destroy plans automatically once dispatched. For a single environment, use `destroy-test.yml` or `destroy-stage.yml` instead; the CLI procedures below remain available for individual-stack cleanup. The separate **PROD destruction** workflow is already protected by the `PROD` environment approval gate; it is not part of this matrix. -### 6.2 Automated destruction of PROD +### 6.2 Manually triggered destruction of PROD 1. Open **Actions → Destroy Infrastructure (PROD)** and choose **Run workflow** on **`main`**. 2. The job waits for required approval in the `PROD` GitHub environment. An authorized reviewer (repository owner, Jørn, or Pau) uses **Review deployments → PROD → Approve and deploy**. This generic GitHub label approves the **destruction job**, not a deployment. @@ -513,9 +552,28 @@ The following workflow outcomes have been reported as successful and can be revi | Independent TEST-only destruction and recreation | **Successful**; STAGE and PROD remained intact | | Independent STAGE-only destruction and recreation | **Successful**; TEST and PROD remained intact | -A screenshot of a full successful run end to end with destruction will be added once STAGE and PROD destruction has been completed. -For an optional workload check, identifying a frontend floating IP from OpenTofu outputs or OpenStack and verify the Nginx default page responds over HTTP port 80 will be given. The assignment permits application-level testing to be omitted. +### Execution Evidence + +The following screenshots document the successful CI/CD pipeline execution, infrastructure destruction, and subsequent recreation of all three environments. + +| Screenshot | Description | +| --- | --- | +| [1. CI Success](screenshots/1_ci_success.png) | Successful formatting, validation, TFLint, and Checkov checks | +| [2. Deployment Success](screenshots/2_deploy_success.png) | Successful deployment of TEST, STAGE, and PROD | +| [3. Destroy PROD](screenshots/3_destroy_prod_success.png) | Successful destruction of PROD infrastructure | +| [4. Destroy STAGE](screenshots/4_destroy_stage_success.png) | Successful destruction of STAGE infrastructure | +| [5. Destroy TEST](screenshots/5_destroy_test_success.png) | Successful destruction of TEST infrastructure | +| [6. Redeployment Success](screenshots/6_deploy_after_destroy_success.png) | Successful recreation of all three environments after destruction | + +#### Successful Deployment + +![Successful deployment](screenshots/2_deploy_success.png) + +#### Successful Redeployment After Destruction + +![Successful redeployment after destruction](screenshots/6_deploy_after_destroy_success.png) + ## 9. Known limitations and improvements @@ -527,3 +585,5 @@ For an optional workload check, identifying a frontend floating IP from OpenTofu 6. **Runner and kOps dependencies are external:** self-hosted runner service availability, kOps cluster, Kubernetes namespace, access control, and instructor permissions require separate setup and ongoing maintenance. 7. **Bootstrap state is not remotely configured in the provided code:** its state needs separate backup and lifecycle protection. 8. **CI/CD tests are infrastructure-focused:** end-to-end application or Nginx availability tests are not implemented; the assignment permits workload-specific tests to be omitted. +9. **Runner bootstrap configuration drift:** The issue was discovered when my public IP address changed, requiring an update to `admin_cidr` to restore SSH access to the runner VM. During the subsequent `tofu plan`, OpenTofu proposed replacing both the SSH security group rule and the runner VM due to a detected difference in its cloud-init `user_data`. NTNU upgraded the SkyHiGh OpenStack platform from Epoxy to Gazpacho during November 4–6, 2026, which may have contributed to infrastructure changes, although a connection to the detected drift has not been confirmed. To avoid unintentionally replacing the runner, only the SSH security group rule was updated using a targeted plan and apply. The runner remains operational, but the cloud-init difference should be investigated before running an unrestricted `tofu apply` in `bootstrap/`. +10. **Documentation is maintained manually:** The README and module documentation require manual updates when infrastructure configurations or workflows change. A future improvement could integrate `terraform-docs` into the CI pipeline to automatically generate and validate documentation for OpenTofu module inputs, outputs, and requirements. This would help keep technical documentation synchronized with the infrastructure code and reduce maintenance effort. \ No newline at end of file diff --git a/screenshots/1_ci_success.png b/screenshots/1_ci_success.png new file mode 100644 index 0000000..0cc742e Binary files /dev/null and b/screenshots/1_ci_success.png differ diff --git a/screenshots/2_deploy_success.png b/screenshots/2_deploy_success.png new file mode 100644 index 0000000..536f93b Binary files /dev/null and b/screenshots/2_deploy_success.png differ diff --git a/screenshots/3_destroy_prod_success.png b/screenshots/3_destroy_prod_success.png new file mode 100644 index 0000000..00ca46a Binary files /dev/null and b/screenshots/3_destroy_prod_success.png differ diff --git a/screenshots/4_destroy_stage_success.png b/screenshots/4_destroy_stage_success.png new file mode 100644 index 0000000..5a4d7ac Binary files /dev/null and b/screenshots/4_destroy_stage_success.png differ diff --git a/screenshots/5_destroy_test_success.png b/screenshots/5_destroy_test_success.png new file mode 100644 index 0000000..76b0fa5 Binary files /dev/null and b/screenshots/5_destroy_test_success.png differ diff --git a/screenshots/6_deploy_after_destroy_success.png b/screenshots/6_deploy_after_destroy_success.png new file mode 100644 index 0000000..3831ae4 Binary files /dev/null and b/screenshots/6_deploy_after_destroy_success.png differ diff --git a/screenshots/runner.png b/screenshots/runner.png new file mode 100644 index 0000000..6e89cb5 Binary files /dev/null and b/screenshots/runner.png differ