What is the "True Cloud"?
An interesting analogy I heard recently via Marc Soester, Managing Director at IPMO, compared the Cloud to a modern day power station. When Thomas Edison established the first public power station in 1882, it enabled the enterprise corporations and the 'well to do' to gain greater access to electricity. The power industry rapidly developed until electricity was more widely available from the 1920s onwards, business could now more effectively consume a mass produced resource rather than having the cost and complexity of generating themselves. Today, this concept is accepted as the normal standard in the generation and distribution of power, a company still investing in its own power generation is a very rare and specialised situation.
In many ways the Cloud is similar, the building blocks pioneered by Amazon Web Services (AWS) in 2002 paved the way for rapid evolution, the industry has already hit a level of maturity where the benefits enabled are not limited to the enterprise corporations and well to do, but are widely available to all. So why is adoption still under 10% in most established markets? It is my opinion that confusion and misunderstanding of what the Cloud actually is plays a major role, hence my use of the term 'True Cloud' within the title of this post.
Firstly, the term "Cloud" is the new buzz word in technology, hence is used very loosely and often attached to a poor cousin of what the true Cloud is. This generates both a false understanding and experience. The 'true Cloud' is a global public power station equivalent, providing a scale and sophistication never before seen. True Cloud investment is in Billions and not Millions, providing security, performance, flexibility and cost effectiveness that is unparalleled. If a Cloud provider is trying to lock you into long term contract early, likely this is not true Cloud but the rental space on a smaller scale data centre - what is often called co-location. I will try to explain further utilising the key benefits of true Cloud.
Is absolute ground zero for AWS and Azure (Microsoft's Cloud platform), with no required lock in contracts, if the security is not up to scratch, customers can simply take their business elsewhere (by also taking advantage of the flexibility true Cloud offers, being the ease to move workloads if there are benefits to do so). Jeff Smith, the CEO of Suncorp Australia when they announced a transition of all available workloads to AWS, publicity stated the following "They have very advanced automation that allows us to move workloads through their data centre and have a lot better security and scalability than anyone we have seen and a lot better than we can do ourselves”. This aligns to my main point, security in the true Cloud is actually better than what anyone else can offer. To further highlight this point, the CIA is a client of AWS in the US and as you would expect, the requirements they place on the security of their data is extremely high. No development within the true Cloud is bespoke, meaning that if any client makes a specific request, such as the CIA, its merits are reviewed across the broader customer base requirements. Therefore the environment established to meet the security needs of the CIA is also available to every other customer of AWS, being a top 4 banking institution or the local coffee shop.
By controlling all the variables, products generally work more effectively, hence efficiently in the true Cloud. A common experience we find is the need to turn off capacity when established to meet a software product's specifications. It is not uncommon to find applications running at under 50% capacity once migrated, hence after monitoring we simply turn off the unused capacity. This is due to the software supplier needing to specify to the lowest common denominator, as the quality of the platform the software will be hosted on cannot be guaranteed. AWS & Azure control every aspect of the environment, from every hardware component to the ideal air temperature, meaning everything within the true Cloud operates at optimum.
This is not only the case for today but into the future. I would think most have all experienced purchasing a new TV only to see an improved and cheaper model available months later. This was also the common experience when having to acquire infrastructure hardware that generally lasts 5 years. Within the true Cloud your workloads are always running on the latest and greatest hardware, providing the performance and cost benefits that come with it.
For some clients today, AWS is the preferred option, others Azure, others still a combination of the two with a hybrid on premise component. With continual evolution within the AWS & Azure Cloud building blocks (solution & service offerings) plus price competition, the reality is that the preferred option can change, even on a daily basis. The ability to move workloads enables the ideal option to always have the ability to be chosen, providing a very different concept to the management of IT infrastructure that takes time for many to get their head around. By making it difficult for workloads to be moved either via lock in contracts or complex environments, providers are removing one of the key benefits that a true Cloud solution should enable. Ever cursed yourself for locking into a provider who didn’t deliver on their promises? Well this can now be avoided.
The traditional approach to infrastructure hardware is to acquire, given the standard lifecycle is five years, a forward estimate would be applied plus some buffer just in case, then the requirements specified and the unit ordered. Not ignoring the Capex versus Opex argument, the simple reality is that the investment tied up in the gap between utilisation and capacity is dead money. With the continuous improvement of processing capacity as well as the evolution of the working environment (such as the Internet Of Things), often the forecasted capacity including the buffer is quickly fully utilised, providing challenges ranging from poor performance (likely to impact customer satisfaction) to unbudgeted additional investments and inefficient bolt-ons. This is on top of the additional cost benefits achieved that optimal performance and flexibility enable.
Finally, when discussing local data centres with CEO's from large organisations, it was proposed to me that wouldn't an AWS or Azure simply absorb these centres as local capacity requires. I don't see this as likely to happen, my expectation is that as sovereignty concerns subside over the coming period, hosting workloads in international locations will become the norm. Some Australian businesses are already taking advantage of cost benefits of hosting in the US based on the current environment. The true Cloud is a truly global solution and AWS & Azure are the leaders today and will be into the future, hence no business needs to choose anything other than global best practice again.