Object Storage Is Becoming the Hidden Contract Between Your Services
Object storage is often introduced simply as a cheap place to put files. But in many data systems, it quietly becomes something much more important: the actual contract between your services.
Imagine a typical architecture: an API starts a batch job that processes data and writes Parquet files, another service reads a JSON summary to serve a dashboard, and a user downloads a CSV export. None of those components talk directly to each other, but they all agree on one thing: where the artifacts live and what they mean.
That agreement is a data contract.
In this post, I will explore why object storage ends up filling this role, how to design paths as APIs, the problems this approach inevitably presents, and what mature systems evolve to use instead.
Why not put everything in the database?
Relational databases are excellent for business state. They are the right place for users,...
Copyright of this story solely belongs to hackernoon.com. To see the full text click HERE
