Performance Test-Performance Test Requirement Analysis
Here, an e-commerce shopping (B2C) website is used as an example:
- Customers' shopping website performance testing (business) requirements:
- From the end of late December to the end of the lunar calendar year (the beginning of February in the next year) (<= 50 days) the estimated website turnover (4 million). Here the turnover can be understood as the total amount of purchase orders completed by the website;
- Access order conversion rate: 10%, which is interpreted as the actual website order;
- Daily access time: 24 hours x 80%. It is understood that normal users will perform e-shopping before eighteen o'clock P.M. AM to AM, and am to am is the peak shopping period;
- Average number of purchased items for each order: About 3 items totaling about 300 yuan. Here, the average consumption of each order is 300 yuan, and the average purchase of 3 items is 3;
- Order efficiency: 85%. This refers to the person who placed the order. If payment is made and the transaction is completed, 15% will cancel the order or no payment;
- Average page views (IP/PV): 10 times. After the IP addresses of all clients accessing the website are counted, each IP address browses 10 webpages on average;
- Business Performance Requirement Analysis: (assuming that all orders occur during peak hours, if each user browses and compares five items, one item will be selected, and each item needs to be browsed by 5 different links on the webpage)
- Average number of times per hour during peak hours: Sales/total sales time/peak time/single order amount/order efficiency = 4000000/50/4/300/0.85 ≈ 8
- Average access volume per hour during peak hours (Order Access volume/order conversion rate): 8*5*5*3/10% = 6000
- Peak traffic = 1.3 x average traffic: 6000x1.3 = 7800
- Number of Client IP addresses accessed during peak periods: 7800/10 = 780
- Design of stress testing scenarios (rough estimation for your reference and understanding only ):
- Design two groups of test scripts, group A generates orders, and group B only browses product-related;
- The number of script executions in group A is 80, and the number of script executions in Group B is 700;
- In the peak test, 100 people submitted orders and 1000 people accessed them online;