Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/stamper/yao-benchmark
Performance comparison test
https://github.com/stamper/yao-benchmark
bun deno django elysiajs expressjs fastapi fastifyjs flask koajs mojo nodejs python
Last synced: about 1 month ago
JSON representation
Performance comparison test
- Host: GitHub
- URL: https://github.com/stamper/yao-benchmark
- Owner: Stamper
- License: unlicense
- Created: 2024-02-08T08:07:18.000Z (10 months ago)
- Default Branch: master
- Last Pushed: 2024-02-12T05:44:24.000Z (9 months ago)
- Last Synced: 2024-10-15T17:30:51.492Z (about 1 month ago)
- Topics: bun, deno, django, elysiajs, expressjs, fastapi, fastifyjs, flask, koajs, mojo, nodejs, python
- Language: TypeScript
- Homepage:
- Size: 39.1 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# yao-benchmark
Yet another one benchmark (that nobody asked for)
### Long story short
Results table
Best result for each framework powered by 1 worker.
| Framework | Runtime | Server | RPM [^1] | L [^2] | CPU% [^3] | Mem% [^3] |
|-------------|---------|----------|----------|--------|-----------|-----------|
| Flask-RESTX | py310 | Gunicorn | xxx | xx | 30 | 10 |
The full data set is presented in this [google spreadsheet](https://docs.google.com/spreadsheets/d/1Cojv3-i-Jp5OBe26qyfy-kO59fkgCiZMujKfK0i-G1U/edit?usp=sharing).
*All the boring details can be found below.*
[^1]: Requests per minute measured by Locust
[^2]: Number of lines of code excluding initial bootstrap (like `startproject` in Django)
[^3]: CPU and Memory consumption measured by `htop` tool during the test
### Server
- Dell Inspiron 3511 / Core i3-1115G4 / 32GB / SSD
- xUbuntu 22.04.3 LTS
#### Tools
```
pyenv 2.3.35
nvm 0.39.7
```
#### Runtimes
```
Python:
- 3.10.13
- 3.12.1
Pypy 3.10-7.3.15
Mojo 0.7.0
NodeJS:
- 18.19.0
- 20.11.0
Deno 1.40.3
Bun 1.0.26
```
#### Python servers
```
Gunicorn 21.2.0
Uvicorn 0.27.0.post1
Hypercorn 0.16.0
```
#### Python frameworks
```
Flask 3.0.2
Django 5.0.2
Starlette 0.36.3
```
#### Python REST-API
```
Flask-RESTX 1.3.0
Django REST framework 3.14.0
Django Ninja 1.1.0
FastAPI 0.109.2
Blacksheep 2.0.6
```
#### JS/TS frameworks
```
Express 4.18.2
Koa 2.15.0
Fastify 4.26.0
Deno 1.40.30
Elysia 1.0.50
```
#### Mojo framework
```
Lightbug 0.1.1-alpha
```
### Testing environment
```
Dell Latitude 5440 / Core i7-1365U / 32GB / SSD
Ubuntu 22.04.3 LTS
Python 3.10.12
Locust 2.22.0 (10 Workers, 100 users)
```
### Network
```
TP-Link TL-WR940N
```
### Testing scenario
```
Request: POST /api/
{
"payload": ""
}
Response: 200 OK
{
"result": ""
}
```
### Locust configuration
```
$ cd locust && locust --master
$ cd locust && locust --worker # 10 times
```
### Conclusion
I really appreciate you've got this long. No conclusions, the numbers speaks for itself.