Automation and Self-Service
In cloud, automation of routine tasks is a defining characteristic. In any hybrid cloud scenario, you should be able to automate the deployment of new resources, such as virtual machines, networks, and storage resources. Even if you’re not a DevOps-drinking developer, you can still adopt some of the infrastructure-as-code methodologies that are bringing new possibilities to the application workload environment. For example, you can create self-service systems that enable individual business units to provision their own resources. Or, consider this scenario: suppose you’re working at a college and want to make sure every new student gets their own virtual desktop and dedicated storage space. You can write a routine that creates these resources automatically each time someone in the Admissions department accepts a new student and changes their record in the database.
If you haven’t enabled deep automation in your data center, you don’t have a hybrid cloud. Remember, self-service is one of the virtues of the cloud; without automation, you can’t introduce self-service into the environment. This doesn’t necessarily mean that every user in your company needs to be able to provision resources, but it does mean that the ability to use resources should not be limited to just the operations staff. In a DevOps world, this capability should extend to, at a minimum the development staff.