We run two AWS accounts, dev and prod to develop our pipelines. Dev has rights to read prod but can only write to dev - this way we can read prod data but develop any data objects in dev that we want to alter. I would like to know how best to deploy lakefs in such a situation? Would you have any docs that might assist with this dual account situation?

Hey @George Pongracz, welcome. I would like to understand your use-case better. Do you intend to use lakeFS for dev data, or production data?

@Yoni Augarten we would like to use it for production data, but we do all of our development in our dev account. For example let's say we have a two existing data objects in prod and want to create a new third one fed from the two existing ones. As part of this we would need to alter one of the two existing ones as well - lets say to get an additional field to be used in feeding the new third object. I am still reading through your documentation :)

Generally speaking, you may want to deploy lakeFS in one of the accounts or both, depending on which data you would like to enable versioning for. If I understand correctly, you would like to have git-like capabilities on your dev environment, while also reading data from the prod environment. Is my understanding of the use-case accurate? To support you in the deployment process, we would like to understand the exact use-case. You are welcome to describe a typical workflow, with as much details as possible (you can PM me if you prefer).