In the post-millennial era of real world cloud deployment, the modern digitally distributed nature of businesses requires a range of infrastructure options to allow each customer to leverage a mix of cloud technologies to best suit their unique needs while optimizing the associated costs. How can we enable this kind of flexibility in the face of infrastructure complexity and still accommodate for a degree of legacy IT in the modern IT stack where operating models are still being reinvented and built for next-generation use cases? Rob Tribe, Vice President System Engineering EMEA, Nutanix explores the answer.
Back in the day, consumers took what they were given. During the post-war years of the 50s, people were largely happy to be able to engage in increasingly modernized societies where choice was limited, but still variable enough to convey some sense of freedom.
Throughout the latter part of last millennium, enterprise buyers and consumers alike started to want things (products, services and everything in between) provided the way they liked them, more closely attuned to personal preferences and requirements. Ultimately, people and organizations both realized, understood and came to appreciate the fact that they could have things the way they wanted.
The power of personalization and the rise of Customer Relationship Management (CRM) created a new type of buyer. Whether we’re thinking about consumer shoppers or enterprise procurement managers, the freedom to choose means a lot more now than it ever has before.
Purchasing power preference penchant
If we combine this penchant for purchasing power preference with the realities of modern digital society structures and think about how cloud computing now forms the underpinning substrate of all business and commerce, we can understand why every cloud needs to be different.
Some cloud deployments will be optimized for Input/Output transactional girth, some will be pumped for CPU speed, some will have an outstanding cadence to perform analytics calls, others will be optimized for storage capacity and so on. We are building these instances of cloud across on-premises private-public hybrid structures in multi-cloud deployments with some poly-cloud complexity further twisting the plot.
Over and above the specific types of cloud that are deployed, every organization will have a different position on the path to cloud evolution. Although there are many cloud-native organizations, the vast estates of legacy IT out there also need to be engineered-in to the modern fabric of the current IT stack.
Hopefully, the personalization parallel is obvious by this point, organizations need their cloud provisioned, served (deployed) and delivered (maintained) in the most flexibly hybrid way possible, with the maximum possible power to consume that IT resource in right way for the business.
Chaos in the cloud
Amidst this need for flexibility and choice, there is also a sort of ensuing chaos in cloud creation, some of which is down to separate cloud environments with their own different management interfaces. We know that Cloud Services Providers (CSPs) naturally differentiate their offerings; after all, it’s part of the way they go-to-market and create a brand. But these differentiations rarely add to any notion of a cohesive multi-cloud bond.
Add those factors to non-negotiable contracts and other limiting factors such as reserve instance contracts and the cloud landscape starts to look harder to consume.
If we take our examination back inside the enterprise itself, we can see that cloud resources (compute, storage, analytics, networking and so on) are often spread across complex disparate team structures. When all these different teams, divisions and departments use different software tolls, portals and data exchange conduits, we get complex cloud, often made more complex.
A unified management plane
If we accept that this ingredient mix forms much of the common cloud deployment landscape, then we can appreciate how a unified management plane can provide a much-needed clear view. From this higher power elevation, we can leverage the same technologies across different lower-level cloud substrates to achieve built-in networking and the ability to use existing security setups.
Suddenly we start to move from a crowded marketplace towards the place we feel we need to be at as buyers. A unified management plane doesn’t just support multiple use cases, it also offers centralized management and governance across the board, no matter which user group (or groups) are accessing its resources.
Let’s remember, everybody needs cloud to be deployed and run in a different way. Developers need cloud services to work for rapid deployment and testing; business managers need cloud services to exhibit highly available access with simplified use case dynamics, operations need cloud to be secure, visible and robust, while also being easy to maintain and update.
No one cloud-fits-all
Looking at the state of cloud today, we know that analyst firm Gartner forecasts end-user spending on public cloud services to grow 23.1% by the end of 2021 to total US$332.3bn, up from US$270bn in 2020. With the rise of spending on different cloud consumption models, from Software-as-a-service (SaaS), Infrastructure-as-a-service (IaaS), and Desktop-as-a-service (DaaS), to self-hosted public clouds, there’s no longer a one-size-fits-all approach.
If there were a single ‘meal’ resulting from cloud computing deployment, then pretty much everyone would want it flavoured differently, served in a different size, formed into a particular shape or mould and delivered on a different plate… and then some people would want a remotely accessible takeaway order too.
Let’s also remember that with the exception of the small companies that were ‘born in the cloud’, today’s businesses rely on a considerable number of existing software and applications that cannot simply be set aside in favour of cloud-native alternatives. Technology suppliers must remain empathetic to their customers’ needs to maintain an existing estate of applications.
Sunsetting legacy applications are costly and takes a lot of time and effort, but organizations shouldn’t be held back by these limitations. Furthermore, the cost of ownership associated with public cloud platforms, private cloud platforms, hybrid (public and private) cloud platforms and traditional/non-cloud enterprise infrastructure can vary immensely, so it’s vital to understand how different deployment models align with their existing workloads and business needs.
Cloud on your terms
Choice in the contemporary cloud doesn’t just mean the ability to spin-up and spin-down the resource capacity of deployed cloud instances – although it does mean that too – it also essentially means the ability to select the ideal hardware, hypervisor and clouds needed for any given workload or wider task process.
If a business can say it has that power over its cloud consumption, then it can say it has cloud on its own terms. Freeing an organization from the anchoring effects of legacy technologies and the drag and slowdown effect caused by licence portability restrictions, organizations can use the cloud they want on any hardware in any formation.
Adopting a progressive approach to a hyperconverged infrastructure (HCI) software solution, enterprises can achieve the personalization they seek from a truly effective hybrid cloud model. A hybrid multi-cloud platform deployment should have the flexibility, simplicity and cost-efficiency needed to run applications in private or multiple public clouds.
All of this happens in a deployment environment where cloud operators and administrators can use a single management plane to truly manage both their private and public cloud infrastructure. A new tier of flexibility is reached as organizations can then manage and operate their cloud estate as one single cloud that they can derive more agility from.
- Why small businesses would benefit from the cloud
- Thinking beyond the cloud
- Organizations neglect DevOps security in latest Cloud Threat Report
- 3 top tips to ensure compliance in the cloud
Cloud was always supposed to convey an inherent sense of flexibility and flexibility is synonymous with choice. Given the many varieties of virtual data storage and transmission that have been developed inside the cloud computing model itself, choice must be the guiding principle behind any workable cost-effective and productive hybrid multi-cloud solution.