Deployment Guide

[jacobbeck-totango]

I'm trying to figure out how to best deploy SuperDuper. There isn't anything in the docs and I don't see anything when searching git.

I'd assume that I could create a dockerfile and deploy via kubernetes or ECS. The docs do mention Ray--which I haven't explored yet--but there isn't any more information on it than a brief mention and a link.

Has anyone actually deployed superduper to production infra? If so, would love some pointers on what you did and why.

[blythed]

Hi @jacobbeck-totango, please watch this space. We are currently developing a slightly different strategy around deployment. We are building a base Cluster object which will allow developers decide how they deploy compute, change data capture, caching, vector-search and job-management. Developers are free to create their own and also contribute implementations. An alternative is our enterprise on-premise ray+kubernetes deployment. Read more about that here: https://superduper.io/.

[jacobbeck-totango]

Thanks @blythed!

Will definitely keep an eye out. We are trying to make a decision quickly for an AI project that I think this repo aligns really well with. We are super cost conscious so trying to find solutions that allow us to self-host easily.

Thanks again!
Jacob