Monitor and Analyze

Learning Objectives

After completing this unit, you’ll be able to:

Monitor performance during and after tests.
Explain how to analyze test results to identify and resolve performance issues.

You’ve done some great work and now the focus is on monitoring and analyzing the results of your test. Monitoring and analysis are crucial for identifying performance issues and optimizing your application. A great motto to keep in mind is, “Optimize, test, then optimize again!”

Define Why Monitoring and Analysis Are Important

Issue identification: Monitoring helps you identify performance issues as they occur, allowing you to take immediate action.
Optimization: Analysis helps you understand the root causes of performance issues and make informed decisions to optimize your application.
Continuous improvement: By monitoring and analyzing your tests, you can continuously improve your application's performance and reliability.

The Test Runs window in Scale Test that shows you completed tests.

Steps to Monitor and Analyze Performance

Effective analysis relies on centralized, real-time insights during execution. This allows you to pinpoint bottlenecks and root causes.

Set Up Monitoring

Monitoring tools: Use monitoring tools to track performance metrics such as response time, throughput, and error rates.
Alerts: Set up alerts to notify you of any performance issues as they occur.

Analyze Results

Live test monitoring with Scale Test: Use Scale Test to view runtime metrics as they happen, including DMLs, platform events, RPS graphs, and Apex hot spots. Pay close attention to the Apex Performance Section for optimization opportunities, potentially using ApexGuru.
Identify Issues: Analyze the results of the test to identify any performance issues or configuration problems.
Focus on percentiles: When analyzing performance, prioritize P90/P95 response times over simple averages. This approach better represents the experience of the vast majority of your users, not just the average case.
Use Scale Center as the source of truth: Monitor key metrics such as governor limits, API latency, database contention, and expensive SOQL queries. The Org Performance page in Scale Center is your go-to for identifying bottlenecks and action items. Use in-depth diagnostics like the consolidated report in Scale Center to investigate the root cause of performance issues. Pinpoint bottlenecks and areas for optimization.
AI-powered insights: Use AI-powered code insights (ApexGuru) to detect inefficiencies in Apex code and SOQL/DML usage, offering targeted recommendations to optimize performance.

Implement Improvements

Address issues: Address any issues identified during the test, making necessary adjustments to your application or test setup.
Continuous improvement: Use the insights gained from your tests to continuously improve your application's performance and reliability.

Use Case: Scale Center Investigation

After running your full-scale test, you use Scale Center to analyze bottlenecks in the checkout and chatbot handoff flows. You create and analyze an investigation in Scale Center. You discover that post-insert Apex logic on the Case object causes CPU time spikes. ApexGuru flags inefficient nested loops in trigger logic. You refactor the code and rerun the test to confirm latency and EPT improvements across high-volume paths.

Key Takeaways

Scale testing requires more than just simulating load, it demands a disciplined, production-informed approach that involves these best practices.

Use production intelligence to define hot spots.
Model realistic user journeys with controlled workloads.
Validate environments with trial runs and baseline comparisons.
Automate test data creation and cleanup for consistency.
Centralize analysis in Scale Center and prioritize percentile-based metrics.
Expand beyond load tests to endurance and stress scenarios.

By embedding these practices into your process, you ensure not only technical readiness but also business confidence in your application’s ability to perform under real-world conditions.

예상 시간

주제

도움말 검색