Files
llm-automation-docs-and-rem…/.gitea/workflows/README.md
dnviti d6d44270ee
Some checks failed
Build / Build and Push Docker Images (worker) (push) Waiting to run
Deploy Staging / Deploy to Staging (push) Waiting to run
Lint / Lint Code (push) Has been cancelled
Security / Security Scanning (push) Has been cancelled
Build / Build and Push Docker Images (api) (push) Has started running
Build / Build and Push Docker Images (chat) (push) Has started running
Build / Build and Push Docker Images (frontend) (push) Has started running
Test / Run Tests (push) Has been cancelled
refactor: reorganize CI/CD pipelines into separate workflow files
BREAKING CHANGE: Monolithic ci.yml split into focused pipeline files

## Pipeline Reorganization

Split single CI/CD pipeline into 7 specialized workflows:

1. **lint.yml** - Code quality checks (Black, Ruff, MyPy)
2. **test.yml** - Test suite with coverage reporting
3. **security.yml** - Security scanning (Bandit)
4. **build.yml** - Docker image builds and registry push
5. **deploy-staging.yml** - Staging environment deployment
6. **deploy-production.yml** - Production deployment (tags only)
7. **docs-generation.yml** - Scheduled documentation generation

## Benefits

- **Modularity**: Each pipeline has single responsibility
- **Performance**: Workflows run independently, faster feedback
- **Clarity**: Easier to understand and maintain
- **Flexibility**: Trigger pipelines independently
- **Debugging**: Isolated failures easier to diagnose

## Dockerfile Improvements

- Fix FROM AS casing (was 'as', now 'AS') in all Dockerfiles
- Resolves Docker build warnings
- Improves consistency across build files

## Documentation

- Added .gitea/workflows/README.md with:
  - Workflow descriptions and triggers
  - Dependency diagram
  - Environment variables reference
  - Troubleshooting guide
  - Best practices

## Migration Notes

- Old monolithic pipeline backed up as ci.yml.old
- All triggers preserved from original pipeline
- No changes to build or deploy logic
- Same environment variables and secrets required

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-10-21 13:37:21 +02:00

175 lines
5.1 KiB
Markdown

# CI/CD Workflows
This directory contains separated CI/CD pipeline workflows for the Datacenter Documentation System.
## Workflow Organization
Each workflow is designed for a specific purpose and can run independently:
### 1. Code Quality (`lint.yml`)
**Triggers**: Push/PR to main/develop
Runs code quality checks:
- **Black**: Code formatting verification
- **Ruff**: Linting and code quality
- **MyPy**: Static type checking
**Purpose**: Ensure code quality standards before merge.
---
### 2. Testing (`test.yml`)
**Triggers**: Push/PR to main/develop
Runs test suite with coverage:
- Unit tests with pytest
- Integration tests (when available)
- Coverage reporting to CodeCov
- MongoDB service container for database tests
**Purpose**: Verify functionality and prevent regressions.
---
### 3. Security (`security.yml`)
**Triggers**: Push/PR to main/develop, Weekly schedule (Monday)
Security scanning:
- **Bandit**: Python code security analysis
- **Safety**: Dependency vulnerability scanning (future)
**Purpose**: Identify security vulnerabilities early.
---
### 4. Build (`build.yml`)
**Triggers**: Push to main, Version tags
Builds and pushes Docker images:
- **Components**: API, Chat, Worker, Frontend
- **Registry**: Gitea Container Registry
- **Caching**: BuildKit cache for faster builds
- **Tagging**: Branch names, SHAs, semantic versions
**Purpose**: Create deployable container images.
---
### 5. Deploy - Staging (`deploy-staging.yml`)
**Triggers**: Push to main (after successful build)
Deploys to staging environment:
- **Platform**: Kubernetes
- **Namespace**: datacenter-docs
- **Components**: API, Chat, Worker
- **Validation**: Rollout status checks
**Purpose**: Test changes in staging before production.
---
### 6. Deploy - Production (`deploy-production.yml`)
**Triggers**: Version tags (v*)
Deploys to production environment:
- **Platform**: Kubernetes
- **Namespace**: datacenter-docs
- **Components**: API, Chat, Worker
- **Validation**: Rollout status + smoke tests
**Purpose**: Deploy stable versions to production.
---
### 7. Documentation Generation (`docs-generation.yml`)
**Triggers**: Schedule (every 6 hours), Manual
Automated documentation generation:
- Collects infrastructure data
- Generates documentation via LLM
- Uploads artifacts
- Commits changes (if any)
**Purpose**: Keep documentation up-to-date automatically.
---
## Workflow Dependencies
```
┌─────────────┐
│ lint.yml │ ─┐
└─────────────┘ │
┌─────────────┐ │
│ test.yml │ ─┤
└─────────────┘ │
├──> ┌──────────────┐ ┌────────────────────┐
┌─────────────┐ │ │ build.yml │ ──> │ deploy-staging.yml │
│security.yml │ ─┘ └──────────────┘ └────────────────────┘
└─────────────┘ │
│ (on tags)
┌──────────────────────┐
│deploy-production.yml │
└──────────────────────┘
┌──────────────────────┐
│docs-generation.yml │ (independent, scheduled)
└──────────────────────┘
```
## Environment Variables
### Required Secrets
- `USERNAME`: Registry username
- `TOKEN`: Registry access token
- `LLM_API_KEY`: LLM provider API key
- `LLM_BASE_URL`: LLM provider base URL
- `MCP_SERVER_URL`: Infrastructure API endpoint
- `KUBE_CONFIG_STAGING`: Staging Kubernetes config
- `KUBE_CONFIG_PRODUCTION`: Production Kubernetes config
### Required Variables
- `PACKAGES_REGISTRY`: Container registry URL (e.g., `git.commandware.com`)
## Running Workflows
### Automatic Triggers
All workflows run automatically based on their configured triggers.
### Manual Triggers
Some workflows support manual triggering via Gitea Actions UI:
- `docs-generation.yml` (workflow_dispatch)
## Best Practices
1. **Always pass lint and test** before merging to main
2. **Review security reports** weekly
3. **Test in staging** before creating version tags
4. **Use semantic versioning** for production releases (v1.2.3)
5. **Monitor deployment rollouts** in Kubernetes
## Troubleshooting
### Build Failures
- Check Docker build logs
- Verify poetry.lock is committed
- Ensure all dependencies are available
### Test Failures
- Review test output in job logs
- Check MongoDB service health
- Verify environment variables
### Deployment Failures
- Check Kubernetes cluster status
- Verify secrets are configured
- Review deployment rollout logs
## Maintenance
- Review and update workflow configurations quarterly
- Update action versions when new releases are available
- Monitor workflow execution times and optimize as needed