So, when it comes to API rate limiting, my approach is pretty straightforward.

dealing with api rate limiting is getting more complex as our services scale. we're considering a few approaches like token buckets and leaky buckets, but also looking at distributed solutions. curious to hear what others have found effective. what trade-offs did you make, and what issues have you run into?