Cloud adoption is set to become a computing norm – even for companies that have until now rejected these technologies. But as hosted software (G Suite, Office 365, Salseforce.com etc) gather pace, most have been unable to completely avoid cloud services.
Much of the discussion around cloud migration suggests that it is a ‘big bang’, all-or-nothing play, with the whole data centre being shifted to the cloud. Although possible (in theory) not all workloads belong in the cloud.
Cloud migration doesn’t have to be ‘big bang’
Many cloud operators give the impression that adoption is not only inevitable but that all of your systems will eventually be hosted. And the sooner this transition takes place, the better.
The reality is that this is your business and your infrastructure, and you are fully justified in moving at your own pace. For various reasons (unfamiliarity, security concerns, uncertainty etc) you have resisted major cloud adoption projects – so it makes sense to maintain a cautious roll-out.
Cloud migration on your terms
One way to maintain control of the process and the speed at which you move is to bring cloud technologies into your data centre first. Using platforms like VMware vCloud Suite, Microsoft Hyper-V virtualisation or OpenStack.
Deploying a private cloud allows your business to migrate applications and workloads, learning how the concepts and technologies apply to your business. At the same time, you can take advantage of automation and self-service to accelerate your IT operations to deliver a better quality of service to your in-house users.
This approach can be more expensive than going with one of the large platforms like AWS, Azure or Google Cloud. With cloud in-house, however, you retain full control of the process so you can migrate applications and servers at your own pace. This makes the transition more manageable and lays the groundwork for when you do decide to migrate to a lower-cost public cloud provider.
Re-engineering workloads for the cloud
One of the key benefits of cloud platforms is their elastic pricing model – you only pay for what you use. However, simply moving your virtual servers into the cloud is not efficient.
Your on-premise systems are configured to run 24x7x365 because there is no reason to let them spin down. But in the cloud where every resource is billable – CPU cycles, RAM, storage etc – you pay for running servers, even when they are not being accessed.
The major cloud platforms allow you to set servers to spin down automatically overnight for instance, helping to reduce costs. However, these servers are themselves considered relatively heavyweight.
The future of operating in the cloud lies in containerisation. This technology breaks applications into blocks that can be created and destroyed automatically according to demand. Unlike a virtual server, the container is a much smaller package, containing nothing but your application code and the libraries required to run it; there is no operating system or additional applications, helping to minimise the number of resources used – and therefore costs.
With a private cloud, you can begin the process of re-engineering and optimising for the cloud before moving to a public cloud platform. This will help to contain costs when you do finally migrate and simplify the process of transition.
To learn more about the moving to the cloud and how to simplify the transition to the cloud, please get in touch.