2 ms·
Show HN: Django-xbench – slow endpoint aggregation for Django
- yeongbin05 7mo agoPreviously it focused on per-request timing. This update adds rolling-window aggregation to detect performance trends.
- simonw 7mo agoGathering the stats in-memory is neat - I'd expected this would be writing to a DB but that's not how it works: https://github.com/yeongbin05/django-xbench/blob/f63316126b5f714e4bc23f6ba511c6f59df41f6c/src/django_xbench/slowagg/window.py#L43-L60 https://github.com/yeongbin05/django-xbench/blob/f63316126b5...
- blorenz 7mo agoTerrific project. I use Strawberry GraphQL so about all of my endpoints will be /graphql unfortunately. It would be nice if it could parse off a header to qualify which bucket it goes to. That way I could send over the operation and django-xbench takes care of it. I see you are going have N+1s on the roadmap. I currently use django-zeal which does a fantastic job detecting them but there isn't a visualization. I'm keenly watching where you go with this. I need more django perf in my life.
- patrick91 7mo agohow are you finding Strawberry?
- yeongbin05 7mo agoThanks for the thoughtful feedback! I haven't planned GraphQL-specific grouping yet. N+1 visualization is the current priority but I'll definitely keep this in mind. Thank you very much