I've been hearing increasing chatter about an (arguably still mythical) trend towards NoOps. From my perspective, NoOps is the idealization that infrastructure can become so abstracted and automated that no dedicated Ops team is required to manage it.
Previously we've seen NoOps scale down for services like Heroku or Digital Ocean, but now products like AWS Lambda and Microsoft Azure Functions have shown that managed compute services can provide infinite scaling in both directions, and do so with little-to-no Ops involvement.
On the other hand, even with services like Amazon S3, Amazon DynamoDB, and Amazon Aurora, we're still not seeing the same level of NoOps-through-managed-services on the data side of the house. And, even worse, it doesn't seem like anyone's even talking about NoOps for the network layer.
Is the end-solution for NoOps ultimately another tier of managed services? Or is the ultimate NoOps tool an automation and abstraction tool that rides above the primatives offered by IaaS providers?