In one of my past projects, I led a non-functional testing stream and had to plan performance tests for 70 applications. My testing team consisted of 8 experienced performance test experts. However, the go-live date was in six months.
Initially, I thought we would never finalize all the performance validations within the available timeline. Due to the high number of affected applications and the limited testing time, planning, implementing, and executing performance tests for all 70 applications was impossible. Our little chance was a risk-based performance engineering approach.
Performance Risk Assessment
First, we created a non-functional-requirement questionnaire, which captured performance-requirement-related topics such as response time, expected transaction volume, and growth estimates. Then, we sent it with a risk assessment sheet to our product teams.
We used these impact criteria
Employee
Customer
Reputation
Revenue
We used these likelihood criteria
Volume
Response Time
Technology
Core Business
Performance Risk Calculation
Second, we used each category's performance risk rating and specified performance test activities. For example, we decided that only performance tests for applications with very high and high performance risk categories would be conducted.
Performance Risk Mitigation
Surprisingly, just 10 percent of our 70 applications fall into the performance test-relevant category. Very liberated from the enormous burden, we prepared and implemented the required performance tests for these ten applications and quickly met the demanding timeline.
Benefits of Risk-based Performance Engineering
1.)
Risk-based performance engineering has helped us to focus our efforts on performance-critical applications and brought transparency to our performance engineering activities.
2.)
We applied this performance risk classification to all our applications. Therefore, any change to services or applications rated in the very high or high category had to be validated by a performance test onwards.
3.)
We finalized our performance testing in time and fixed several critical issues.
4.)
Everyone understood the importance of performance engineering.
5.)
We no longer talked about finding a budget for performance engineering because, due to the performance risk classification, it was a planned action.
Keep up the great work! Happy Performance Engineering!