| Criteria | HTTP | Testing with Real Browser |
|---|---|---|
| User simulation | No client-side rendering | Real user simulation |
| Script implementation and customization | Difficult when web sites are complex | Simple scripts, easy to customize using EveryStep recorder |
| Script replay | Low-level analysis required | You see what you get using EveryStep |
| Script maintainability | Programming skills required | Easy, as almost all done by EveryStep Recorder |
| Multi Browser Support | Can emulate web browser using Agent-string | Chrome and Edge |
| Footprint on the load injection machine | Low, up to 500-1000 sessions per load injector | High, up to 8-25 sessions per load injector |
| Recommended for DevOps | Depends on the actual test scenario | Yes, easy to use and realistic figures |
| Recommended for Load Tests | • Webservers • Web APIs: JSON, SOAP, Rest, XML. • Single web pages without JS or CSS |
• Web Applications • Complex pages containing Java Scripts and CSS |
| Pro’s and Con‘s | Pro‘s Low Cost Exact HTTP/S targetCon‘s No scripting No JS or CSS |
Pro’s Real User Simulation Get what you see in the browser exactly Support for JS, CSS, RIACon’s High Cost Lot‘s of data |
| Costs | Low | High |