The virtualization trend of recent years has improved the use of existing resources – without making substantive changes to the underlying procedural model. In most cases the only real issue was the cost of procuring the hardware; the auxiliary expenses (electricity, licenses, support etc.) were rarely broken down to the level of the individual machines and frequently dismissed as ‘always incurred anyway’. But the emergence of the cloud and its transparent cost structure are necessitating the implementation of equivalent clarity within the company’s own organization. Otherwise it is impossible to compare prices between different providers. Besides virtual machines, there are other services, for instance databases, that can be used as well. The benefit of these ‘databases as a service’ is that an actual operating system is not necessary, putting an end to maintaining the machines or the operating systems themselves.
Fear of the unknown
What happens then in the IT community is often a natural, emotional response to sweeping change. Won’t the employees be let go if IT is no longer responsible for basic operations? Might my own job in the company be at risk? These and other thoughts provoke defensiveness. Arguments such as the need to keep data in-house for security reasons are primarily put forward to avoid having to address the issue. And in adopting this approach, many companies forget that the pertinent question is not whether these things will be relocated to the cloud, instead quite simply when.
Smart use of the cloud and its significant benefit
If we move another step forward in the transformation process to address the costs of operating a server, it is perfectly possible that the hourly rate for running one instance in an extremely efficient data center will be lower than when booking a cloud provider. Scenarios like this arep robably quite common. But they do not consider a fact that is among the biggest advantages of cloud infrastructure: Pay per use. The companies only pay for what they actually consume. Naturally, the price can frequently seem high if you base the calculation on 24/7. But there are completely different options available if the resources are deployed smartly and only when they are actually needed. For instance, dynamic operation during off-peak hours – so from 6:00 pm to 9:00 am – can bring significant savings, as it enables a switch to a weaker instance.
These and similar scenarios need to be taken into consideration to realize the full potential savings. Local ‘gear’, in contrast, does not offer anything like as much cost-optimized flexibility.
Going one step further
Infrastructure as a service is only the start of the journey. Companies can rightly claim to have grasped the issue of cloud computing when they substitute the operating system with platform as a service and use innovative plans like function as a service, in which only a small fee is ultimately payable when a function is accessed.
Virtualization of server systems has become self-evident for many companies. The flexibility of using resources more efficiently in data centers – and saving money in the process – is a step that almost everyone is taking. But the same companies are evidently having trouble embracing the next stage of evolution, namely migrating the virtual machines to the cloud. And the reasons are more than just technical.
The now widespread corporate austerity policies have rendered the area of innovation essentially non-existent, as it does not contribute to the primary value chain. What is it good for anyway? Servers were scaled, purchased, operated for several years with the necessary software, and then replaced with a more powerful model. This mantra worked without a hitch for decades. But progress is forcing companies to rethink their strategies, and is putting the fear of god into many of the IT employees.
Infrastructure and Apps, Digital Transformation & Innovation