Public Cloud providers have long evangelised the sustainability benefits of their platforms. The environmental elements of these claims are largely based on their energy efficiency, specifically their Power Usage Effectiveness (PUE), their use of renewable energy and “economies of scale”. Whether you buy into these ideas or not, the imperfections of PUE as a comparative metric and the upcoming swathe of sustainability legislation, directives, and frameworks (EED, CRSD, TCFD, SECR, ESRS just to start) mean that things are about to get very real for the large hosting, colocation and public cloud providers. But more than that, to achieve our sustainability goals, organisations need to be agile and adaptable while at the same time reducing risk and cost.
Additionally, with increased scrutiny of competition in the cloud computing market, solutions that simplify moving workloads between clouds (i.e. hybrid multicloud solutions) mean it is increasingly the smart approach for the medium to long term. To help better appreciate this, here are three reasons why hybrid multicloud is the sustainable choice.
Sharing Resources (Cloud Bursting)
Organisations often have large amounts of IT resources sitting idle, usually for redundancy or meeting infrequent peak requirements. Even if those resources are then turned off, an organisation is still responsible for the embodied carbon (Scope 3 emissions) associated with the manufacture and transportation of hardware, which has a clock running against its useful lifespan (support life). So being able to easily make use of shared infrastructure, like public cloud, can be a great way to help mitigate this embedded environmental (carbon footprint) impact.
There are lots of examples of use cases where the public cloud works very well for this kind of approach. For example, test environments, LLM generation, batch processing, seasonal scaling, analytics, and anything where a lot of computing is required for a relatively short amount of time. However, as per a recent case study, disaster recovery (DR) can provide one of the clearest business cases.
The traditional approach to DR has been to run a second facility with dedicated resources standing idle, waiting for a disaster or test event. The “Near Zero infrastructure DR” approach in the public cloud can eliminate the need for a permanent on-premise secondary site by reducing the ongoing DR footprint to a small “pilot light” in the public cloud. Resources are then scaled up as required, usually only for a few days or weeks per year of DR test activity, and then scaled back down.
Eliminating the need for a duplicate second site should (theoretically at least) reduce the data centre footprint by up to 50%, so it’s a pretty clear-cut case from a sustainability and cost perspective. In one respect, it's a bit like only renting a secondary accommodation when you go on vacation or have a problem with your regular home, rather than buying a second property. Additionally, whilst you aren’t using the scale-up resource, someone else can be. This means that the embodied emissions of the underlying hardware can potentially be shared between a number of different organisations if the cloud/hosting provider is able to provide data on its attributed emissions.
Maybe this approach doesn’t fit the risk profile of every application at every organisation, but for the majority of less critical applications it could provide one the biggest sustainability wins and a significant cost savings potential. Having a core platform that easily enables hybrid cloud can mean that IT teams can start to make use of these strategies, as well as look for other ways they can temporarily shift workloads to make use of renting in the public cloud vs buying up front for the data centre.
It’s pretty well established that if we (humans) are to meet our sustainability goals, then technology and digital transformation will play a big, if not decisive, part. But if the benefits are to be fully realised, then digital transformation also needs to reach an organisation's most critical applications and data.
Moving to the cloud has sometimes been billed as a “transformational project” in itself, with “cloud first” often being the mantra. However, experience has shown IT leaders that simply migrating to the public cloud doesn’t mean you’re automatically getting a transformation. After all, the cloud is just someone else's data centre, a point proven (as if it needed to be) by the fact that most CSP revenue still comes from IaaS and storage services.
IT budgets have poured into these “lift n’ shift” migrations, yet what remains on-premises often languishes from neglect and underinvestment. Ironically, workloads stay put often because of their critical nature and data sovereignty requirements. This leaves the most valuable data, primed for transformative potential, stranded.
Having an effective hybrid cloud infrastructure that unifies services and operations across public and private clouds can unlock innovation in the data centre by facilitating the movement of applications and data whilst maintaining a common operating model. Policy-based management means that data replication can be easily managed to avoid potentially sensitive information reaching undesired locations without the proper transformation. This can give teams the freedom to innovate on the hardest-to-reach applications whilst still maintaining the privacy of their data.
Additionally, if the task is to just move as quickly as possible to the public cloud, then a hybrid cloud model that provides the same data centre class services across public and private clouds can help accelerate the migration. Consistent operations between on-premises and public cloud means that the lengthy application transformation, rearchitecting and refactoring can be reduced or potentially even completely avoided.
So whether it’s grabbing the low hanging fruit or the stubborn, data centre dependent applications, a hybrid cloud can provide the on ramp to accelerated transformation across an organisations IT estate whilst potentially providing some more immediate emissions reductions.
Risk Mitigated by Choice
As per the shared resource (cloud bursting) example discussed above, reducing the cost of DR and improving the business continuity posture is one way that the hybrid cloud can address risk. Interestingly, this approach can also address some climate risks, such as data centre failure due to excessive heat or water stress, as well as the commercial or technical risks associated with vendor lock-in.
As recently as summer 2022, after days of unusually hot weather, several UK data centres suffered issues ranging from partial to complete outages. In one catastrophic case, an entire hospital had to resort to paper processes for several weeks because of their poorly implemented BC/DR approach. However, these types of issues are not limited to the on-prem data centre world.
Even the biggest public cloud providers are not immune. Besides the climate risk, you may recall a time at the start of the pandemic when, in certain regions, demand for some IaaS services exceeded capacity as IT teams rushed to set users up to work from home. This unexpected widespread spike in demand meant that some cloud admins looking to provision new resources were met with an error reporting lack of resource availability, starkly ending the myth that the cloud is of infinite capacity.
Supply chain issues haven’t necessarily disappeared with the pandemic. In an uncertain world, and with 2024 predicted to be the biggest election year in history, keeping your options open and being able to choose where your applications and data live could make the difference between doing business and not. Or, if that seems a touch dramatic, consider the commercial implications of vendor lock-in and what happens when a technology provider gets too sticky. Choice doesn’t just doesn’t mean the ability to adapt to an ever-changing world, but it also provides the leverage to get the best deal and the best service from your suppliers, including how they meet your wider sustainability requirements.
Looking to the future
If you look at the marketing, you’ll notice that the big public clouds are happy to discuss hybrid cloud, but pretty much allergic to the word “multicloud” and for fairly obvious reasons; it implies revenue easily moving away from their platform to a competitor's cloud.
If nothing else, hybrid multicloud means choice. The choice to move workloads between hardware platforms, vendors, clouds and hypervisors puts CIOs and their teams in control of their infrastructure vs being at the mercy of big tech and their sales teams that are sometimes times their customers’ size. It means the choice to put workloads in the right place based on technical, commercial and environmental considerations or whatever the business demand derives.
Choice puts the customers in the driving seat for the long term, and that’s not just good for individual organisations but also good for the market and even the planet as a whole.