# Test Orchestration
Parallel test execution architecture and test suite organization.
**Main orchestrator:** `.github/workflows/test-parallel.yaml`
## Performance
- Wall-clock time: 10-15 minutes
- 50% faster than sequential
- Up to 10 parallel runners
## Execution Flow
```
load-versions (1 min)
↓
build-adapters + build-docker (parallel, 8-10 min)
↓
test suites (parallel): shared, cli, mcp, core, launch, frameworks
↓
summary (1 min)
```
## Test Suites
| Suite | Purpose | Artifacts | Time |
| -------------- | ------------------------------------ | ----------------- | ------ |
| **shared** | Language-agnostic debug fundamentals | Adapters + Docker | 5 min |
| **cli** | dev-cli commands and services | None | 4 min |
| **mcp** | MCP server and tool handlers | Adapters + Docker | 6 min |
| **core** | Core API, DAP client, session mgmt | Adapters | 5 min |
| **launch** | VS Code launch.json integration | Adapters + Docker | 4 min |
| **common** | Common utilities (path, config, env) | None | 2 min |
| **logging** | Structured logging framework | None | 2 min |
| **frameworks** | Flask, Express, Spring Boot, etc. | Adapters + Docker | varies |
## Language-Grouped Framework Tests
Frameworks grouped by language (matches local dev-cli execution):
- **Python:** django, flask, fastapi, pytest
- **JavaScript:** express, jest
- **Java:** junit, springboot
Each runs: `./dev-cli test run -s frameworks -l {language}`
## Matrix Strategy
Language-specific suites use matrix for DRY:
```yaml
test-shared:
strategy:
matrix:
language: [python, javascript, java]
```
Jobs display as: `test-shared (python)`, `test-shared (javascript)`, etc.
## Debug Logging
All workflows support `debug_logging` input:
```yaml
env:
AIDB_LOG_LEVEL: ${{ (inputs.debug_logging == true) && 'TRACE' || 'INFO' }}
AIDB_ADAPTER_TRACE: ${{ (inputs.debug_logging == true) && '1' || '0' }}
```
Trigger via: `gh workflow run test-parallel.yaml -f debug_logging=true`
## Adding New Suites
1. Add to dev-cli `src/aidb_cli/services/test/suites.py`
1. Add job to `test-parallel.yaml`:
```yaml
test-new-suite:
needs: [load-versions, build-adapters]
uses: ./.github/workflows/test-suite.yaml
with:
suite: new-suite
suite-name: New Suite Description
python-version: ${{ needs.load-versions.outputs.python-version }}
needs-adapters: true
```
1. Add to summary job `needs` array
## Caching
- **pip/npm/maven:** Cached via setup actions
- **Docker:** Layer caching via buildx + GitHub Actions cache
- **Adapters:** Keyed by `versions.json` hash
## Auto-Detecting Summary
Script `.github/scripts/format_job_summary.py` parses `${{ toJson(needs) }}` to auto-detect results. Generates markdown table, exits with failure if any job failed.
## References
- `.github/workflows/test-parallel.yaml` - Main orchestrator
- `.github/workflows/test-suite.yaml` - Reusable test runner
- `.github/workflows/test-frameworks.yaml` - Framework tests
- `docs/developer-guide/ci-cd.md` - Complete CI/CD reference
