About

To assist in demonstration and testing a basic Python project that executes multiple requests against a OpenWebUI API to generate a number of requests simulating a number of requests against a ChatGPT application. In the initial release a single sample message is submitted multiple times over and synchronously. To simplify use the project is containerized and allows for simply starting the container to execute a number of requests. Should multiple instances be desired the container can be started multiple times to increase the load against the API.

Note

This is for experimentation, learning and demonstration purposes ONLY and is not intended for large scale test or validation for production implementations

Built With

• 
• 

Getting Started

To simplify usage, simply clone the repo to your local development machine and build the container and run.

Requirements

To simplify distribution and use, it is desired to be ran using a container runtime such as Docker or Podman. Instructions to build the container is provided for both Docker or Podman installations.

Tested with

• Podman v1.16.0
• Docker version 27.3.1

Steps

1. Clone Repo to your local development machine

   git clone https://github.com/rjohnston6/chatgpt-load-gen.git
   

2. Define environment variables for target OpenWebUI Target URL and API Key and desired base request count. A env_example.list file is included rename the file to env.list and update the variables to set them as part of the container execution process.

   Variable	Default Value
   COUNT	10
   API_KEY	YOURSUPERSECUREKEY
   API_URL	http://localhost:8000/v1

3. Build the the container using your preferred container runtime

   • Docker:

     cd chatgpt-load-gen
     docker build -t chatgpt-load:v1.0 .
     

   • Podman:

     cd chatgpt-load-gen
     podman build -t chatgpt-load:v1.0 .
     

4. Start the continer to generate load.

   • Docker:

     docker run -d --rm --env-file env.list chatgpt-load:v1.0
     

   • Podman:

     podman run -d --rm --env-file env.list chatgpt-load:v1.0
     

Roadmap

• Add randomized list of prompts for requests made