Cloud not beyond obsolescence
11 April 2016 | 0
At the most recent TechFire event, mention was made of a most interesting term: legacy cloud infrastructure.
Take a moment to let that sink in.
The reference was made by Chris Ducker, head of proposition marketing Europe, Sungard Availability Services during the panel discussion. It was in the context of maturing use of cloud services and the development of hybrid IT environments. In these hybrid use patterns, some applications and capabilities are kept in-house while others are either placed in or consumed from the cloud. The event explored the fact that the new normal for most organisations would be some variation, from one end of the spectrum to another, of this type of usage. The balance would depend on various factors from the nature of the business and sector, to the available skills and the technology outlook of the organisation.
Ducker was making the point that as cloud computing usage matures, it will be inevitable that at some point, certain usages or models will become obsolete, hence the term legacy cloud infrastructure.
“With the mass economics of the cloud as a significant factor for any cloud usage, an older arrangement might start to become hard to ignore when those itemised bills come in every month”
These kinds of instances may have been early cloud deployments of applications, or usage of services which no longer conform to the best of what cloud has to offer and so, much in the same way that any bit of tech, tin or kit will eventually become older than a bag of old hat, the same will happen with cloud.
Many early deployments of applications in the cloud were done on the basis of a single, or small set of needs, mostly around capacity or availability, which may have been difficult or impossible to accomplish in-house or on-premises. As such, the cloud developed around the container, virtual machine, platform or whatever receptacle was employed, at a furious pace and so that application may have now be regarded as legacy cloud infrastructure.
This is most intriguing, as most of the indicators with which we associated the evaluation of a server, application or service, do not apply to cloud. For example, is it able to scale? Probably. Does it still perform well? More than likely. Is it highly available, secure and resilient? Yes. That’s what cloud does. And yet, it may still be considered legacy as new connectors, application programming interfaces (API) or other widgets, might now be available around the original deployment or service that are far beyond its capability and so render it obsolete nonetheless.
The good news is that dealing with legacy cloud infrastructure and deployments should be far easier than their physical counterparts of old. Firstly, an entirely new environment could be created in short time, with even more power and capability, and tested extensively before the critical deployment need ever be troubled. Secondly, migration should be nothing more than a controlled failover, the execution of which the users might never even notice. And finally, the entire process — should catastrophe strike — will be 100% reversible, unlike going back to those large hunks of out-of-support tin and switchgear that may well be worth more in scrap or nostalgia value than they are as working equipment.
However, that very luxury may well lead to a certain lack of motivation to switch away from legacy cloud infrastructure, as it doesn’t creak, it doesn’t fall over and it doesn’t complain. It may, however, become expensive and inflexible. These criteria might then become the best drivers to modernise, in all its software-facilitated ease. With the mass economics of the cloud as a significant factor for any cloud usage, an older arrangement might start to become hard to ignore when those itemised bills come in every month. If that is coupled with constant questions from production, marketing, HR or payroll along the lines of “why can’t our cloud do x?” then the more conventional indicators of the need to upgrade may become irrelevant.
However, while contemplating this development in obsolescence, it also set me thinking about the other end of the spectrum. With those oft quoted examples of Uber and Air BnB as great disruptors — to which I would add another participant in the recent event, MicksGarage.com — there is still room for innovation despite the near ubiquity of cloud capabilities, for almost every pocket.
It would appear as if even small businesses have the ability to leverage the ever evolving capabilities of cloud infrastructure and services to innovate. Some do this, like the autoparts seller above, by putting their critical infrastructure in the cloud and keeping special IP and ‘secret sauce’ development in-house where it can be tightly controlled and exercised, and some by doing the opposite, and many more by finding some balance in between that suits their business model, processes, industry, season or day of the week. The point is that businesses seem to be coming round to the idea that the key feature of ‘the cloud’ is its flexibility to allow organisations to achieve their vision. The very malleability and flexibility of these ever more sophisticated services has allowed business models that would never have succeeded in the purely or at least more heavily, physical infrastructure world — whether these are retailers that provide stock market-like pricing on retail web sites, or businesses that have such sophisticated analytics that they can release updates to their primary platform for real-time user interaction in lieu of more traditional user testing for near instant identification, and critically, rectification of issues.
So while the spectre of legacy cloud is growing and will loom over many even as the bleeding edge momentum is washing over the latecomers, there still seems to be room for innovation, tweaking and fine-tuning for innovative usages that will likely continue to disrupt more and more businesses.
While there seems to be no end to the number of businesses taking to the cloud, there also seems to be no end to the innovative use of this evolving resource. Far from the prospect of early twentieth century worry of having to close patent offices because every significant invention had already been made, the prospect for cloud seems to be a steady but swift pace of development that will see the term legacy cloud infrastructure become as much of our everyday ICT parlance as innovation, disruption and start-up.