I love the idea of using Kubernetes operators to create and manage a project's infrastructure. Like the PostgreSQL operator or some examples of API gateway operators.
But something tells me that this can be a problem. Like in the past, when we used to store logic inside databases, using triggers and stored procedures. And now we know this kind of architecture can lead to problems related to performance, scale, and vendor lock-in.
I know that Kubernetes isn't attached to a single vendor, as we can use it without AWS, Google, or other dependencies. But can we be repeating some error from our past broadly using Operators?
