Life, in general, is seldom linear. So are people’s minds and the decisions we tend to make. There are days when we go through the ebbs and lows of decision making. It can feel clouded at times. What is critical though (albeit tricky), is to try being objective, avoid being impulsive, and feel the need to join the rat race. To be rational whilst embarking on a major decision.
The world of networking is no different. Digital transformation and cloud is the buzzword, and rightfully so. However, in that milieu, organizations seem to be in a bit of a quandary in terms of optimizing their infrastructure. The decision – ‘to cloud ‘ or ‘not to cloud’- looms large.. There is indeed a cloud of uncertainty (pun intended).
Enterprises have traditionally invested in traditional datacenter networks. In all fairness, it has served its time well but has also seen challenges. From maintenance, lack of elasticity, to higher Cost of ownership (TCO), rising CAPEX and OPEX… the list is long.
The choices that the I&O teams need to make are also evolving. Driven by dynamic business demands, there is a need for IT teams to consider the shift towards cloud; and evaluate the merits rationally. The cloud has presented both exciting opportunities and some significant challenges.
With the growing demand for modernised distributed applications , the networking world is forced to relook and optimize their traditional operating models. From storage, to servers, to software, no area of the datacenter has been spared. In most cases, customized hardware is expensive, particularly when performing a task that can perhaps be easily solved using software.
In simple terms, businesses seek to adopt a multi-cloud architecture in order to address multiple scenarios. From avoiding vendor lock-in, to optimizing between cost and SLAs, while fostering greater agility. Software defined datacenter (SDDC) is on the rise, for it allows for greater flexibility in datacenter configurations and in turn helps reduce costs.
There is little doubt about the benefits of cloud, for it is one of the most significant platform shifts in the history of computing. It has already impacted billions of dollars of IT spend, and continues to do so. The operational and economic value proposition is immense in terms of infrastructure being available immediately, offering flexibility in terms of the operating model thereby allowing companies to focus on new products and growth.
According to a finding by Canalys, Cloud infrastructure services spending increased 36% to US$47.0 billion in Q2 2021, as workload migration and cloud native application development accelerated.
However, the journey to the cloud is easier said than done. The shift in workloads also brings in complexity in managing requirements. It is imperative to evaluate the pros and cons prior to embarking on a multi-cloud strategy such as:
- Operational: Avoiding vendor lock-in, better cost and performance, flexibility, reduced operational burden, security, Posture, governance and compliance.
- Flexibility: Determining which specific workloads benefit from the elasticity, geographical flexibility, or new technological innovation that the cloud offers?
- Adaptable: Determine if the current or planned workloads can embrace lift and shift i.e.; are they architectured well to adopt cloud-native technologies.
- Costs: Determine if the ROI is indeed justified by comparing the cloud adoption model vis-a-vis On-premise model.
- Best Practices: Determine if best practices on cloud governance are being followed and the gap in skills addressed.
While shifting to cloud has its benefits, there are differing views. The debate on the actual benefits of public cloud versus custom built infrastructure rages on. According to the research by Andreessen Horowitz (source: the cost of cloud a trillion dollar paradox), there are examples of companies such as Dropbox and the likes who have seen significant impact on revenues owing to repatriation of workloads from public cloud to a custom infrastructure , and some who have adopted a hybrid approach.
At the end of the day, organizations need to figure out a balanced approach that works best in the context of their business and operational needs. The key is to evaluate the operational model and optimize consumption based on the nature of the workloads, on its merits.
Cliched as it sounds, applications are the soul and lifeline of business. As the wise said “choose what’s good for the soul, not the ego”.