Over the last two weeks, I have fully immersed myself in the world of cloud. Thanks to conference keynotes, breakout sessions, customer briefings, and ad hoc discussions, I have not had much time to not think or talk about cloud. While this is not necessarily a bad thing, I do have to admit, something has been gnawing at me over these past two weeks: There is too much overloaded, confusing, or misunderstood cloud jargon.
The thing is, just about everyone in the cloud space is guilty of misuse at one point in time (I know I make my contributions). However, it is good to cast a light on some of this behavior in hopes that we can fix it and prevent the consumer confusion it can sometimes cause. Therefore, because I want to pitch in and do my part to help correct this behavior, and because lists are fun to make, here is my opinion on the top five overloaded/confusion-causing terms in the cloud space today:
Service: I am not sure there is any term more used, yet less specific than the term service. I suspect this is the case regardless of what technological context the term ‘service’ comes up in, but I know it for a fact when used in the context of cloud discussions. In cloud computing, services can be servers, storage, networking, business processes, application platforms, applications, and much more. In my view, it is certainly fine to position your cloud computing solution as helping users quickly deliver services. The only problem I have is when you say that but do not bother disambiguating what you mean by service. Tell us exactly what ‘services’ you are delivering.
Dynamic: Ask yourself this: Is there anything that is not dynamic when it comes to the cloud? The term has a variety of meanings ranging from autonomic behavior to elasticity of resources. I use the term dynamic from time-to-time, and in many cases, consumers call me on the carpet and ask exactly what I mean. That is fair and is in fact the right thing for them to do due to the overuse and frankly, abuse of the term. I hope that providers are using the term where appropriate and they are ready to explain exactly what they mean by ‘dynamic’.
Workload: Like ‘service’, the term ‘workload’ is not at all specific in the context of cloud, but it comes up all the time. Unlike ‘dynamic’, I don’t see a lot of misuse of this term, but confusion ensues because of one primary reason. The meaning of workload is highly dependent on who you are talking to within an organization. As an example, to an application platform administrator, provisioning a new application platform may be a workload. However, for a developer that deploys applications to that platform, their definition of workload probably involves application requests. Help everyone out here. When you talk about workloads within a cloud computing solution, start by defining the workload.
Virtualization: The confusion I typically observe in this space results from the fact that when users hear ‘virtualization’ they usually think hypervisors and virtual machines. In fact, the concept of virtualization is simply a logical abstraction of a discrete set of resources. Cloud solutions today apply virtualization at the server level, application infrastructure level, and even the application level to name but a few. Like with many of the terms mentioned above, when using ‘virtualization’ or ‘virtualized’, don’t just use it as a sweeping generalization. Be specific. Tell us exactly what you are virtualizing and what that means to the end user.
Hybrid cloud: While there is constant bickering about public and private clouds, many seemingly agree that hybrid clouds are the eventuality of many cloud architectures. The problem is, we overload, and in my opinion, often misuse the term ‘hybrid cloud’. To me, a hybrid cloud architecture is one consisting of both private and public clouds. Connecting your traditionally deployed on-premise application environments with cloud-based environments in the public realm is extremely valuable, but I cannot bring myself to call that a hybrid cloud. Hybrid architecture, sure, but not a hybrid cloud. I think the difference here is important because hybrid clouds and hybrid architectures dictate different usage scenarios. We need to be consistent and accurate when using the term ‘hybrid cloud’.
This is my list of overloaded and sometimes misunderstood terms in the cloud. What do you think?