Your Tool For Python Performance Tracking
Project description
perf8
THIS IS ALPHA QUALITY, UNSUPPORTED, RUN AT YOUR OWN RISKS
Your Tool For Python Performance Tracking
perf8 is a curated list of tools to track the performance of your Python app.
The project is pluggable, and ships with a few tools:
cprofile - a cProfile to Dot graph generator
pyspy - a py-spy speedscope generator
memray - a memory flamegraph generator
psutil - a psutil integration
asyncstats - stats on the asyncio eventloop usage (for async apps)
Installation
Use pip:
pip install perf8
If you use the cprofile plugin, you will need to install Graphviz to get the dot utility. See. https://graphviz.org/download/
Usage
Running the perf8 command against your Python module:
perf8 --all -c /my/python/script.py --option1
Will generate a self-contained HTML report, making it suitable for running it in automation and produce performance artifacts.
You can pick specific plugins. Run perf –help and use the ones you want.
Async applications
Running the asyncstats plugin requires to provide your application event loop.
In order to do this, you need to instrument your application to give perf8 the loop to watch. You can use the enable and disable coroutines:
import perf8
async def my_app():
await perf8.enable(my_loop)
try:
# my code
await run_app()
finally:
await perf8.disable()
To avoid running this code in production you can use the PERF8 environment variable to detect if perf8 is calling your app:
import os
if 'PERF8' in os.environ:
import perf8
async with perf8.measure():
await run_app()
else:
await run_app()
Screencast
0.0.1 - 2023/01/06
Initial release
0.0.0 - 2022/11/11
Don’t use it.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
File details
Details for the file perf8-0.0.1.tar.gz
.
File metadata
- Download URL: perf8-0.0.1.tar.gz
- Upload date:
- Size: 656.6 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.10.7
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 7dadf24869cae705b3027a2349611d40134283cbc5900e34be128489fd688e84 |
|
MD5 | 4bbea92405d17db47f1b8801ec53fb06 |
|
BLAKE2b-256 | ff1d5985968bbca76bedd748eb6a017ba1a2368f925c682eaace3b5454669abb |