Test Fix

Systematic approach to diagnosing and fixing failing tests.

Process

Step 1: Identify Failing Tests

cargo test --all
cargo test test_name -- --exact --nocapture

Step 2: Reproduce Locally

# With debug logging
RUST_LOG=debug cargo test test_name

# Force single-threaded for race conditions
cargo test test_name -- --test-threads=1

Step 3: Diagnose Root Cause

Step 4: Verify Fix

# Run multiple times
for i in {1..10}; do cargo test test_name -- --exact || break; done

# Run full suite
cargo test --all

Step 5: Regression Prevention

Add test for the specific bug that was fixed.

Debugging Checklist

• Run failing test in isolation
• Check environment variables
• Review recent changes
• Check for missing .await
• Verify database connections
• Look for race conditions
• Check type compatibility

Tools

# Debug logging
RUST_LOG=debug cargo test

# Full backtrace
RUST_BACKTRACE=full cargo test

# Specific module
RUST_LOG=memory_core=debug cargo test

When to Skip vs Fix

Skip (temporarily):

• External service down
• Platform-specific issue
• Known upstream bug

Fix immediately:

• Logic error
• Incorrect assertion
• Missing error handling