This blog post first appeared as a guest post on the Vidispine blog.
This is a watershed moment. There are no longer any real technical barriers to media production in the cloud, only cultural and organisational. This obviously doesn’t mean that cloud production is easy or cheap, but it does mean that you no longer need a research project usinguntested bleeding edge technologies to make it a reality.
If you are looking at moving more things to the cloud then this blog post provides a shorthand for how to quickly evaluate the maturity of suppliers’ cloud offerings–are they Fledglings or perhaps already Cloud Natives?
The Laggard requires professional services staff from the supplier to meet with you to gather requirements and then go off and make a custom installation on virtual machines in the cloud. Everything is static and the only difference from an on-premise installation is that the servers are in someone else’s data centre.
Anyone who is still here is way behind and unlikely to ever catch up. The product is most likely a legacy application that should either never be deployed in the cloud, or will be replaced by a cloud alternative.
The Fledgling has an application that may be almost as technically adapted to cloud as the Cloud Native, but the infrastructure management and organisation has not yet reached full maturity.
New resources have to be requested via email, or online forms and become available a few hours later at best. This is mostly because someone has to pick up the request manually and provision the service specifically for that request. Anything submitted on Friday afternoon in the supplier’s time zone may not be serviced until Monday morning.
Level: Cloud Native
The Cloud Native can provide dynamic and scalable resources without direct involvement of the supplier. The virtual resources may still be visible to the customer, but they can be requested using self-service portals, or ideally through well-documented APIs.
The Cloud Native offering is unlikely to even be available as an on-premise option, as it relies on cloud infrastructure functions, for example to provide dynamic scaling. Organisationally the supplier may not even be able to provide support for on premise deployments.
Level: Black Box
The Black Box provides the software service without exposing anything about how it is actually delivered. By definition, regular Software-as-a-Service (SaaS) offerings should all be Black Box, but beware that some SaaS offerings do not auto-scale to the required size and hide manual engineering processes behind a slick sales portal.
A type of Black Box that fewer people are familiar with is the serverless or Function-as-a-Service (FaaS) offerings that provide programmatic services and where the machine instances are completely invisible (and irrelevant) to the requester. Examples of these are AWS Lambda and Vidispine’s own VidiNet service.
Serverless services make some people uneasy as much of the complexity is hidden and it will be difficult to understand how the service is delivered (especially engineers like to peek under the hood). This unease is somewhat ironic when voiced by people in media, where black box hardware systems was the norm for many years.
Now to be clear, this post isn’t saying that everyone has to be Black Box and that a Fledgling is automatically a worse option than a Cloud Native. Obviously many more things play into supplier selection. This is simply a shorthand to quickly evaluate the cloud maturity of an organisation and detect if there is perhaps a mismatch between the marketing and the actual delivery capability. Lastly, bear in mind that moving up the maturity ladder takes years–not months.