examens-arbete

Requirements

• Bash Version 4+
• Vegeta
• jq
• Python 3.9 (See requirements.txt per framework)

TODO?

• Websockets
• FastApi request models / no models
• Frontend för att visa datan i d3?

ruff

Ruff is used for python linting

ruff . --fix

Run testing

Testing

Basic usage

./benchmark.sh [flags]

./benchmark.sh -t 30s -u http://localhost:7777 -m only -s '["raw_json"]' 

Usage help

./benchmark.sh -h
Usage: benchmark.sh [flags]
    -h                 Show help menu
    -l                 List all available stages
    -m <mode>          Mode: full, skip, only, low_load (full)
    -r <rate>          Requests per second, 0=inf (0)
    -s <stages>        Stages to include/exclude as JSON array, e.g., '["raw_json", "parse_url", "delay"]' (all)
    -t <duration>      Duration in si units (1s, 5m, 10h)
    -u <url>           Base URL of benchmark (http://localhost:5000)
    -w <max_workers>   Maximum number of workers for vegeta (100)

OS

MacOs

Due to MacOs shiping with bash 3.2 and this program relying on newer bash features. ./benchmark.sh points to a newer version of bash installed with brew.
If running on MacOs do

brew install bash

Other

Remove the first line of the code in ./benchmark.sh.
The first line should be the path to your bash installation example: #!/bin/bash

Vegeta report

To view the results of the benchmark use vegeta report, e.g.

vegeta report result.bin
vegeta report -type=json results.bin > metrics.json
vegeta results.bin plot > plot.html
vegeta report -type="hist[0,100ms,200ms,300ms]" results.bin

FastAPI

FastAPI presented with uvicorn under gunicorn

Setup

Create venv

/path/to/python -m venv ./venv

Activate venv

source ./venv/bin/activate

Install requirements

pip install -r ./requirements.txt

FastAPI server

start development server:

uvicorn main:app --reload

start production server:

gunicorn -k uvicorn.workers.UvicornWorker -w 12 -b 0.0.0.0:7777 main:app

Docker

Create image:

docker build -t fastapi .

Start image:

docker run -d -p 7777:7777 --name fastapi fastapi

Flask

Presented with Gunicorn running gevent workers

Setup

Create venv

/path/to/python -m venv ./venv

Activate venv

source ./venv/bin/activate

Install requirements

pip install -r ./requirements.txt

Flask server

Start development server:

flask

Start production server:

gunicorn -w 8 -b 0.0.0.0:7778 app:app

Docker

Create image:

docker build -t flask .

Start image:

docker run -d -p 7778:7778 --name flask flask

Frontend WIP

Setup

Install dependencies

yarn

Start dev server

yarn dev