Protocol-level scripts are designed to simulate the network traffic that an application generates when communicating with a server. By operating directly at the communication protocol layer (e.g., HTTP, TCP/IP), these scripts bypass the client-side Graphical User Interface (GUI). This allows them to precisely measure the time it takes for a request to travel to the server, be processed, and for the response to return. This measurement is a direct reflection of network latency and server processing time, which together constitute the network response time, without the overhead of client-side rendering.

A. GUI scripts: These measure the total end-to-end user-perceived response time, which includes significant client-side processing and GUI rendering time, making them imprecise for isolating network response time.

B. API scripts: While they operate closer to the network than GUI scripts, they are a specific implementation of protocol-level scripting. "Protocol-level scripts" is the more accurate and encompassing category.

C. HTTP scripts: Similar to API scripts, HTTP scripts are a specific type of protocol-level script. The term "Protocol-level scripts" is the correct general classification for scripts that measure network interactions.

1. ISTQB® Certified Tester Performance Testing Syllabus (Version 2018): Section 3.2.2

"Scripting

" distinguishes between the two main types of scripts: "Protocol-level scripts simulate the network traffic that is generated by the client application when communicating with the server... The response time measured by a protocol-level script does not include the time it takes for the client to render the GUI." This directly supports that protocol-level scripts are used for measuring network-centric timings.

2. Jiang

Z. M.

& Hassan

A. E. (2015). A Survey on Load Testing of Large-Scale Software Systems. IEEE Transactions on Software Engineering

41(11)

1091-1118. This survey discusses load generation techniques

noting that protocol-level request generation is a primary method. It states

"Workload generators can generate requests at the protocol level (e.g.

HTTP requests)..." (Section III-A). This confirms that protocol-level is the fundamental approach for simulating network requests to measure performance.

3. Courtois

J. (2012). Performance Testing Guidance for Web Applications. Microsoft Patterns & Practices. While from a vendor

this is a foundational text. Chapter 5

"Web Application Performance Testing Core Concepts

" explains: "Most performance-testing tools work at the protocol level... This means that the tool itself

rather than a browser

generates the HTTP requests." This highlights that protocol-level interaction is the standard for performance measurement tools.