Load Test Monitoring

Overview

BlueClerk runs automated load tests every Sunday at 12am UTC to ensure the platform can handle real-world traffic patterns under stress. The test simulates 1000 virtual users performing mixed read and write operations, then automatically cleans up all test data. Results are saved with detailed breakdowns of read vs. write performance, helping admins monitor system reliability.

How Load Testing Works

Test Configuration

• 1000 virtual users - Simulates significant concurrent load
• Batched execution - 50 concurrent users per batch to avoid overwhelming the system
• Mixed operations - Realistic blend of read-heavy queries and write operations
• 90-second timeout - Test stops automatically if it takes longer than expected
• Automatic cleanup - All test data is deleted after the test completes

What Gets Tested

Each virtual user performs a realistic workflow:

1. Job list query - Fetches 20 most recent jobs (simulates dashboard view)
2. Job count query - Aggregate query for dashboard stats
3. Company lookup - Searches for companies (simulates search/filter)
4. Write operation (20% of users) - 1 in 5 users creates a system health log entry

This mix reflects real usage: mostly reads with occasional writes.

Test Results

Performance Metrics

After each test completes, BlueClerk saves:

• Total operations - Number of database queries executed
• Read operations - Count of read-only queries (job lists, counts, company lookups)
• Write operations - Count of write queries (health log entries)
• Average response time - Mean latency across all operations
• P95 response time - 95th percentile latency (worst 5% of requests)
• Error count - Number of failed operations
• Test duration - How long the full test took to run

Accessing Results

1. Log in with an admin account
2. Navigate to Admin > Operations > Health
3. Scroll to "Load Test Results" section
4. View historical test runs with pass/fail status

Auto-Cleanup

Test Data Removal

After the test completes successfully:

• All health log entries created during the test are deleted
• Test company data is preserved for the next test run
• No leftover test data pollutes your production database

Safety Measures

• Dedicated test company - All test data uses a specific company ID (load-test-company)
• Isolated operations - Test queries never touch real customer data
• Automatic timeout - Test stops after 90 seconds to prevent runaway processes

Viewing Test History

Load test results appear in the System Health page alongside smoke test results:

• Pass/fail indicator - Green checkmark for successful tests, red X for failures
• Performance summary - Read count, write count, average response time
• Timestamp - When the test ran
• Trend analysis - Compare results over time to spot performance degradation

Why This Matters

Regular load testing helps ensure:

• The platform scales - 1000 concurrent users is realistic peak traffic
• Database performance - Queries remain fast under load
• No regressions - New code doesn't slow down critical operations
• Confidence in reliability - You know the system can handle real-world usage