Files
attune/docs/testing/running-tests.md
2026-02-04 17:46:30 -06:00

441 lines
8.6 KiB
Markdown

# Running Tests - Quick Reference
This guide provides quick commands for running all tests across the Attune project.
**Note:** Attune uses a **schema-per-test architecture** for true test isolation and parallel execution. See [Schema-Per-Test Architecture](./schema-per-test.md) for details.
---
## Quick Start
```bash
# Run all tests (from project root)
make test
# Or run individually by component
make test-common
make test-api
make test-executor
make test-worker
make test-sensor
make test-cli
make test-core-pack
```
---
## By Component
### 1. Common Library
```bash
cd crates/common
cargo test
```
**Coverage**: 539 tests
- Repository tests (all 15 repositories)
- Model validation
- Configuration parsing
- Error handling
**Note**: Tests run in parallel with isolated schemas (no `#[serial]` constraints)
---
### 2. API Service
```bash
cd crates/api
cargo test
```
**Coverage**: 82 tests
- Unit tests (41)
- Integration tests (41)
- Authentication flows
- CRUD operations
**Performance**: ~4-5 seconds (parallel execution with schema isolation)
---
### 3. Executor Service
```bash
cd crates/executor
cargo test
```
**Coverage**: 63 tests
- Unit tests (55)
- Integration tests (8)
- Queue management
- Workflow orchestration
---
### 4. Worker Service
```bash
cd crates/worker
cargo test
```
**Coverage**: 50 tests
- Unit tests (44)
- Security tests (6)
- Action execution
- Dependency isolation
---
### 5. Sensor Service
```bash
cd crates/sensor
cargo test
```
**Coverage**: 27 tests
- Timer sensors
- Interval timers
- Cron timers
- Event generation
---
### 6. CLI Tool
```bash
cd crates/cli
cargo test
```
**Coverage**: 60+ integration tests
- Pack management
- Action execution
- Configuration
- User workflows
---
### 7. Core Pack
```bash
# Bash test runner (fast)
cd packs/core/tests
./run_tests.sh
# Python test suite (comprehensive)
cd packs/core/tests
python3 test_actions.py
# With pytest (recommended)
cd packs/core/tests
pytest test_actions.py -v
```
**Coverage**: 76 tests
- core.echo (7 tests)
- core.noop (8 tests)
- core.sleep (8 tests)
- core.http_request (10 tests)
- File permissions (4 tests)
- YAML validation (optional)
---
## Running Specific Tests
### Rust Tests
```bash
# Run specific test by name
cargo test test_name
# Run tests matching pattern
cargo test pattern
# Run tests in specific module
cargo test module_name::
# Show test output
cargo test -- --nocapture
# Run tests serially (not parallel) - rarely needed with schema-per-test
cargo test -- --test-threads=1
# See verbose output from specific test
cargo test test_name -- --nocapture
```
### Python Tests (Core Pack)
```bash
# Run specific test class
pytest test_actions.py::TestEchoAction -v
# Run specific test method
pytest test_actions.py::TestEchoAction::test_basic_echo -v
# Show output
pytest test_actions.py -v -s
```
---
## Test Requirements
### Rust Tests
**Required**:
- Rust 1.70+
- PostgreSQL (for integration tests)
- RabbitMQ (for integration tests)
**Setup**:
```bash
# Start test dependencies (PostgreSQL)
docker-compose -f docker-compose.test.yaml up -d
# Create test database
createdb -U postgres attune_test
# Run tests (migrations run automatically per test)
cargo test
# Cleanup orphaned test schemas (optional)
./scripts/cleanup-test-schemas.sh
```
**Note**: Each test creates its own isolated schema (`test_<uuid>`), runs migrations, and cleans up automatically.
### Core Pack Tests
**Required**:
- bash
- python3
**Optional**:
- `pytest` - Better test output: `pip install pytest`
- `PyYAML` - YAML validation: `pip install pyyaml`
- `requests` - HTTP tests: `pip install requests>=2.28.0`
---
## Continuous Integration
### GitHub Actions
Tests run automatically on:
- Push to main
- Pull requests
- Manual workflow dispatch
View results: `.github/workflows/test.yml`
---
## Test Coverage
### Current Coverage by Component
| Component | Tests | Status | Coverage |
|-----------|-------|--------|----------|
| Common | 539 | ✅ Passing | ~90% |
| API | 82 | ✅ Passing | ~70% |
| Executor | 63 | ✅ Passing | ~85% |
| Worker | 50 | ✅ Passing | ~80% |
| Sensor | 27 | ✅ Passing | ~75% |
| CLI | 60+ | ✅ Passing | ~70% |
| Core Pack | 76 | ✅ Passing | 100% |
| **Total** | **732+** | **✅ 731+ Passing** | **~40%** |
---
## Troubleshooting
### Tests Fail Due to Database
```bash
# Ensure PostgreSQL is running
docker ps | grep postgres
# Check connection
psql -U postgres -h localhost -c "SELECT 1"
# Cleanup orphaned test schemas
./scripts/cleanup-test-schemas.sh --force
# Check for accumulated schemas
psql postgresql://postgres:postgres@localhost:5432/attune_test -c \
"SELECT COUNT(*) FROM pg_namespace WHERE nspname LIKE 'test_%';"
# If needed, recreate test database
dropdb attune_test
createdb attune_test
```
**Tip**: The schema-per-test approach means you don't need to reset the database between test runs. Each test gets its own isolated schema.
### Tests Fail Due to RabbitMQ
```bash
# Ensure RabbitMQ is running
docker ps | grep rabbitmq
# Check status
rabbitmqctl status
# Reset queues
rabbitmqadmin purge queue name=executor.enforcement
```
### Core Pack Tests Fail
```bash
# Check file permissions
ls -la packs/core/actions/
# Make scripts executable
chmod +x packs/core/actions/*.sh
chmod +x packs/core/actions/*.py
# Install Python dependencies
pip install requests>=2.28.0
```
### Slow Tests
```bash
# Run only fast unit tests (skip integration)
cargo test --lib
# Run specific test suite
cargo test --test integration_test
# Parallel execution (default, recommended with schema-per-test)
cargo test
# Limit parallelism if needed
cargo test -- --test-threads=4
# Serial execution (rarely needed with schema isolation)
cargo test -- --test-threads=1
# Cleanup accumulated schemas if performance degrades
./scripts/cleanup-test-schemas.sh --force
```
**Note**: With schema-per-test isolation, parallel execution is safe and ~4-8x faster than serial execution.
---
## Best Practices
### Before Committing
```bash
# 1. Run all tests
cargo test --all
# 2. Run core pack tests
cd packs/core/tests && ./run_tests.sh
# 3. Check formatting
cargo fmt --check
# 4. Run clippy
cargo clippy -- -D warnings
```
### Writing New Tests
1. **Unit tests**: In the same file as the code
```rust
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_something() {
// Test code
}
}
```
2. **Integration tests**: In `tests/` directory
```rust
// tests/integration_test.rs
use crate::helpers::TestContext;
#[tokio::test]
async fn test_integration() {
// Each test gets isolated schema automatically
let ctx = TestContext::new().await;
// Test code using ctx.pool, ctx.app, etc.
}
```
**Important**: No need for `#[serial]` attribute - schema-per-test provides isolation!
3. **Core Pack tests**: Add to both test runners
- `packs/core/tests/run_tests.sh` for quick tests
- `packs/core/tests/test_actions.py` for comprehensive tests
---
## Performance Benchmarks
Expected test execution times:
| Component | Time | Notes |
|-----------|------|-------|
| Common | ~0.5s | Parallel execution |
| API | ~4-5s | **75% faster** with schema-per-test |
| Executor | ~6s | Parallel with isolation |
| Worker | ~5s | Parallel execution |
| Sensor | ~3s | Parallel timer tests |
| CLI | ~12s | Integration tests |
| Core Pack (bash) | ~20s | Includes HTTP tests |
| Core Pack (python) | ~12s | Unittest suite |
| **Total** | **~60s** | **4-8x speedup** with parallel execution |
**Performance Improvement**: Schema-per-test architecture enables true parallel execution without `#[serial]` constraints, resulting in 75% faster test runs for integration tests.
---
## Resources
- [Schema-Per-Test Architecture](./schema-per-test.md) - **NEW**: Detailed explanation of test isolation
- [Testing Status](testing-status.md) - Detailed coverage analysis
- [Core Pack Tests](../packs/core/tests/README.md) - Core pack testing guide
- [Production Deployment](./production-deployment.md) - Production schema configuration
- [Contributing](../CONTRIBUTING.md) - Development guidelines
## Maintenance
### Cleanup Orphaned Test Schemas
If tests are interrupted (Ctrl+C, crash), schemas may accumulate:
```bash
# Manual cleanup
./scripts/cleanup-test-schemas.sh
# Force cleanup (no confirmation)
./scripts/cleanup-test-schemas.sh --force
# Check schema count
psql postgresql://postgres:postgres@localhost:5432/attune_test -c \
"SELECT COUNT(*) FROM pg_namespace WHERE nspname LIKE 'test_%';"
```
Run cleanup periodically or if you notice performance degradation.
---
**Last Updated**: 2026-01-28
**Maintainer**: Attune Team