In this guest post, Will Grannis, founder and director of the Google Office of the CTO, sets out how adopting a multi-cloud strategy can help enterprises navigate IT challenges around value, risk and cost.
Innovation in the public cloud continues at a staggeringly fast pace. Cloud service providers (CSPs) are continually bringing new services to market, offering a greater depth of choice to enterprise customers. The conventional approach of picking a single cloud vendor and ‘locking-in’ to its infrastructure is effective in the context of such rapid change.
Multi-cloud, however, has emerged as a means for capitalising on the freedom of choice the public cloud offers, not to mention, the differentiated technology each supplier in this space can provide. While the meaning of multi-cloud appears self-evident, in reality it’s more about shifting the role of the organisation’s IT function.
IT needs to shift its focus towards consuming ‘as-a-service’ and having the tools in place to architect across different cloud platforms, and then re-architect over time as business requirements change. In essence, a multi-cloud strategy is a service-first strategy.
Multi-cloud means service-first
For most businesses, this as-a-service approach has become the norm. According to Gartner, by 2021, more than 75% of midsize and large organisations will have adopted a multi-cloud and/or hybrid IT strategy.
The advantages of having a multi-cloud strategy are closely aligned with businesses’ broader IT goals of revenue acceleration, improved agility and time to market, as well as cost reduction. Reduced supplier lock-in, increased scope for cost optimisation, greater resilience and more geographic options all serve to provide a stronger basis for operating applications and workloads at scale.
A greater choice of geographical and virtual locations is important for enterprises that have strict requirements about where data can be stored and processed. The EU’s General Data Protection Regulation (GDPR), for example, included the ‘right to data portability’, specifying that personal (e.g. employee or customer) data must be easily transferred from location to location.
In fact, TechUK’s Cloud2020 Vision Report recommends that organisations ingrain data portability and interoperability into their systems. This means that as more options come online, they are then able to re-evaluate their decisions as necessary.
The aim of a multi-cloud strategy should be to build and maintain capabilities to assess an IT workload and decide where to place it, balancing out a variety of factors. Considerations for placement include how workload location can help IT maximise the value delivered to the business, risks associated with a particular deployment choice, costs of each placement and how well each choice fits with its surrounding architecture.
Multi-cloud as the route to innovation
Adopting multi-cloud means the enterprise is able to embrace the new technologies offered by each CSP it uses. Not only does this create cross-cloud resilience, it also gives the ability to run workloads in specific locations that are not offered by every CSP.
Furthermore, it ensures organisations can take advantage of the latest offerings around emerging technologies like machine learning and the Internet of Things.
This level of flexibility must also be supported by interoperability built into the fabric of the multi-cloud’s design. Using a containerised approach, based on open source technology such as Kubernetes, enterprises can ensure that they are able to fluidly move data between cloud platforms and back into their private environments, in order to make compliance and best-fit service provisioning effortless.
Having this level of interoperability brings significant efficiency benefits. In a single cloud architecture, IT teams needed to be able to write programmes specifically for whichever cloud platform they were using.
Organisations are forced to have several teams of developers with siloed skill sets that cannot be applied to more than one technology stack.
Write once, run anywhere with containers
With containers, they have the cross-platform compatibility to write once, and run anywhere. This is an approach is supported by Google Cloud’s Anthos multi-cloud management platform.
Anthos is an open source based technology, incorporating Kubernetes, Istio, and Knative, that lets teams build and manage applications in existing on-premises environment or in the public cloud of their choice.
As the use of multiple cloud services within a single enterprise IT environment becomes the reality, enabling public and private clouds to work in harmony needs to be an organisational priority.
Finding a right-fit environment for each workload means IT can accelerate application development with performance, cost and resilience at its foundation, helping unlock the true range of possibilities a multi-cloud world has to offer.