Is ‘repatriation’ the Beginning of the End for Public Cloud?

The public cloud providers are continuing to grow at an amazing rate, even though both AWS and Microsoft Azure reported a dip in growth in the most recent quarter.  At the same time, there is an increasing chorus of voices talking about the problem of cost management in the cloud and how cloud can often become much more expensive than originally intended. And then there are the emerging stories of repatriation – workloads that are moving from the public cloud platforms back to on-premises or private cloud environments. Does this signal the end for public cloud?

The benefits of public cloud are very clear, compelling and increasingly relevant in the more fast moving, dynamic and changeable world that we’re becoming. The problem is that there is no ‘one-size-fits-all’ cloud service – not even from AWS or Microsoft Azure. And when the wrong cloud is used for the wrong purpose, organisations will often face challenges with cost, reliability and security. The answer isn’t to slow down, or reverse, use of cloud – rather it is to embrace a multi-cloud strategy so that you’re using the right cloud for the right purpose.

Let’s look at typical scenario. Most organisations are already using cloud for business applications like Office 365, very few are thinking about going back to on-premises Exchange and SharePoint servers.  Many organisations are already using cloud for new applications that are built from the ground-up to be ‘cloud native’ – by this I mean security is baked into the application and it is elastic and dynamic so that it minimises the use (and cost) of the underlying cloud service.  And now, some customers are starting to ‘lift and shift’ workloads out of their ‘legacy’ datacentres onto the same cloud platform that they’ve used to build cloud native apps. And therein lies the problem; traditional applications were not built to operate on a cloud native platform – they’re not dynamic and elastic (and hence don’t minimise its use of the cloud service) and they typically assume a traditional approach to security where risks are mitigated at various layers rather than within the application itself. It is no surprise that it is these usecases that create the challenges that lead to organisations considering repatriation.

But there is another option. Rather than pulling these traditional workloads from a cloud native platform to an on-premises environment – multi-cloud enables the option to move those workloads onto a cloud service that is optimised for traditional architectures.  We call these ‘cloud hosting’ platforms as distinct from ‘cloud native’ platforms and are powered by proven and familiar technology like VMware and Oracle that has powered the legacy platforms in traditional datacentres for decades.

Hence, at UKCloud, we encourage our customers to demand choice & flexibility as part of their cloud strategy – which leads them to multi-cloud. With a multi-cloud strategy they can move traditional workloads onto ‘cloud hosting’ technology stacks, develop new applications on ‘cloud native’ technology stacks and integrate all of these cloud services with ‘non-cloud’ workloads such as AIX, Mainframes or even Supercomputers. Multi-cloud embraces the trend of ‘Distributed Cloud’ such that customers can run the same cloud platform across Public Cloud and Private Cloud (for example in Crown Hosting or on-premises).  In this way, costs are optimised and security is enhanced as you’re using the right cloud service for every workload. And in turn, that avoids the need to ‘repatriate’ workloads that have been moved onto a ‘one-size-fits-all’ hyperscale platform.

Of course, multi-cloud introduces many choices and options which can make it seem impossible to navigate. But fear not, our multi-cloud experts are available to help you understand the art of the possible and create an actionable plan to use the right cloud service for every requirement.

TechMarketView thinks that edge genuinely threatens to undermine a key advantage that super scale cloud service providers (CSPs) – in this instance Amazon Web Services (AWS), Google, IBM and Microsoft – have so far enjoyed in the cloud hosting and analytics market. By decentralising storage and processing capacity to locations closer to data and customers, it gives hundreds if not thousands of tier two cloud service providers, telcos, systems integrators and other suppliers a cheaper, simpler way to crash the party which is faster to deploy and offers a performance advantage for certain types of application, particularly in Internet of Things (IoT) environments.