I’m working on a performance capture library for Python because I often need to know the performance of backend systems I maintain. I frequently build tooling to capture performance and save it for later analysis. I/O operations get costly when writing lots of data to disk and creating good real-time analytics tools takes a lot of my time. I wanted a library that captures real-time performance analytics from Python backends.
https://github.com/jakeroggenbuck/kronicler
This is why I wrote kronicler to record performance metrics while being fast and simple to implement. I built my own columnar database in Rust to capture and analyze these logs.
To capture logs, `import kronicler` and add `@kronicler.capture` as a decorator to functions in Python. It will then start saving performance metrics to the custom database on disk. You can also use the middleware for FastAPI.
You can then view these performance metrics by adding a route to your server called `/logs` where you return `DB.logs()`. You can paste your hosted URL into the settings of usekronicler.com (the online dashboard) and view your data with a couple charts. View the readme or the website for more details for how to do this.
I'm still working on features like concurrency and other overall improvements. I've added a lot since the last time I shared on HN. I would love some feedback to help shape this product into something useful for you all.
Thanks! - Jake