In a recent article on CloudTech.com, Monica Brink tries to shatter the illusion that auto-scaling, or the automatic resizing of application resources depending on demand, is an unachievable feat. She is right. That is, if the leadership of an organization has the mindset and There’s no denying it. No matter how cloud solutions are advertised, there is no end-all solution to the myriad concerns in leveraging the cloud. Companies that wish to migrate to the cloud are at a disadvantage if they believe it is as simple as picking a cloud, a couple tools and voila! you’re in the cloud. If this is the mental approach to the cloud then it’s no wonder users are becoming disillusioned by its delivery.
Even the best tools require some initial training and programming. And more likely than not, the organization itself has to undergo a culture change as well as a reallocation of responsibilities and duties as processes become decentralized, self-service and automated.
Brink’s argument is that in order for a cloud to automatically scale to demand, the platform would have to have the intelligence to understand various causes for volume changes and take appropriate action with no human involvement. She states that auto-scaling is a far more involved process than vendors would have you believe. Again she is right, if you follow that mentality.
Successfully auto-scaling and even fully utilizing the cloud is indeed very involved. The initial steps to choose a public cloud (or a few), equip it with the correct tools and then ensure its proper usage, is a demanding job (which also debunks the myth that the cloud will take jobs away). Auto-scaling can happen, but it requires the IT staff to “teach” systems how to understand when it is time to scale up, shut down or any other necessary function. This can be difficult and quite complex in an enterprise (large or small) with diverse requirements across different teams. Or it can be relatively easy, depending on how well equipped the cloud is with the tools that can react to change intelligently.
With an active policy platform, an IT department can create a strategy defining what processes demand auto-scaling, implement those policies and then from that point forward auto-scaling should take place with relative ease.
Moving to the cloud doesn’t mean you have a self-driving vehicle. The cloud exists to provide agile, on-demand compute capacity with the seamless movement of data at scale. And that takes a little work on our part. But that doesn’t mean it can’t happen. In fact, it is achievable and working for the most innovate companies. This takes time, a solid strategy, and a dedicated team with the vision for what the cloud can do in the long run. It is important for organizations to understand this before undergoing a major infrastructure shift.