Recent CIF research suggested that hybrid cloud is very well established across organisations and is no longer the new paradigm. Multi-cloud is now the biggest growth area for many organisations as they have now come to terms with hybrid cloud models.
Hybrid cloud is a concept that has been in mainstream use for many years now but has invariably referred to organisations adopting an approach of blending public (hyperscale) cloud technology with an on-premise or Service Provider cloud. It does not typically refer to organisations adopting multiple public cloud solutions.
However, the traditional ‘hybrid approach’ is changing with organisations increasingly adopting multiple public cloud vendors, often in addition to their on-premise or Service Provider cloud and so the ‘cloud environment’ is becoming ever more complex.
The increasing prevalence of multi-cloud adoption has even spawned a new term to describe it: polynimbus strategies are those that are explicitly designed to embrace multiple public clouds, software as a service, private clouds and/or Service Provider clouds and are often driven by a desire to avoid dependency on one public cloud provider.
The very nature of a polynimbus strategy means that there is no simple reference architecture. Organisations may use different providers for infrastructure as a service (IaaS) and platform as a service (PaaS), with a third for software as a service (SaaS) or they may balance their IaaS across multiple clouds for resilience and performance.
There are many drivers, including technical flexibility, compliance and governance requirements, a desire to avoid vendor lock-in or simply a desire to adopt best of breed solutions irrespective of vendor.
Whilst all of these drivers are understandable and rational when examined in isolation, a polynimbus strategy can also create significant challenges in terms of management, security, governance and cost management.
As with all technology decisions in the modern organisation, there is a trade-off between the benefits of the polynimbus approach and the associated complexities that adopting multiple clouds inevitably brings.
To deliver on the benefits of a multi-cloud approach, organisations must first become cloud-agnostic. The ‘loyalties’ we have formed to particular vendors need to be set aside because it is only when true agnosticism is achieved will organisations be able to properly create the interoperability that is necessary to enable effective management and governance.
Many organisations are still struggling with building effective management systems for their hybrid cloud (private plus one public) environments. Those that have succeeded have been through a process of evaluation between the public cloud vendors and often gone ‘all in’ with one chosen vendor’s tooling and approach.
Becoming cloud-agnostic and re-visiting these original changes is likely to leave many IT Operations leaders with their heads in their hands groaning ‘not again’.
Why would an organisation seek to create further complications in their environment by risking adding further additional cloud environments to their existing mix? As with any solution, going ‘all in’ with one cloud will have meant compromise. The different public cloud vendors have different strengths and weaknesses and tools and options that are available on one may not be available elsewhere.
The main public cloud vendors have many points of convergence but, equally, they also diverge in a number of areas and this isn’t going to change in the foreseeable future.
There are also issues of governance that cause a level of nervousness within some organisations. Some cloud adoption has been tentative because of an inherent fear of passing all the infrastructure to a single provider and many polynimbus strategies are developed to avoid this. They provide an exit strategy or a balanced risk approach and can accelerate pubic cloud adoption in some of the more reticent sectors.
Whatever the driver, one thing is certain. A polynimbus strategy will create management issues unless these are baked into the initial strategy. Having an approach that allows a single view of the IT estate irrespective of its actual location has always been a challenge for organisations with cost and complete often outweighing benefit which, in turn, leads to compromise when selecting appropriate execution venues for different elements of IT.
One example is cost management. If an application is split between multiple clouds then tracking the cost of the application as an entity can be challenging. Organisations need mechanisms that allow costs from multiple sources to be tracked back to their internal structures. Planning this upfront and tagging different elements as being part of an internal application or structure is a vital first step but one that is often overlooked.
Another common issue is monitoring. We have traditionally monitored IT from the infrastructure up to ensure everything is working but where the infrastructure is spread across multiple venues we need a new approach. We need to begin our monitoring from the application side and adopt a top-down approach, monitoring the user experience and using intelligent tooling to identify where issues exist.
In essence, we need to change our thought processes and approach and then we can begin to reap the rewards of being able to choose whichever execution venue is relevant and not be constrained by previous decisions to select one cloud provider over another.
When we break free of that constraint, then the currently reported cloud adoption statistics are likely to increase, our options increase and our risk of compromised functionality and compliance constraints will vanish.
The key to a successful polynimbus strategy is to ensure that it is a strategy and not an accident. Rather than trying to resist the approach and then having to pick up the pieces when the business acts independently, now is the time for IT to embrace it.
Article by channel:
Everything you need to know about Digital Transformation
The best articles, news and events direct to your inbox