cool man. I only took a quick look but one question I had was if this runs as a service - since you say to point clients at it - then would it not make more sense to cache on the node the client is talking to s3-accelerator on since that is the first hop? Genuinely curious. I've been working on my own s3/blob project for the past year but it's purpose is a bit different than yours but does involve caching for reads and I'm just curious if you had a reason for that decision that I hadn't considered and might change how I view my own project https://github.com/afreidah/s3-orchestrator (we are both so creative at naming lol). Threw a star on yours so I remember to go back and look at it in more detail later.