We are both privileged and challenged to live in a time when the pace of innovation is not only blistering, but accelerating. Cloud computing has gone mainstream and is causing upheaval across the IT landscape. The growth in spending on cloud computing, as well as the ecosystem of software and services that is growing around it, paints a picture that predicts the accelerated pace of innovation will continue for some time to come. However, this disruption means that enterprises with a need for agility should consider adopting a multi-cloud approach.
Over the past decade, I've helped companies innovate in IT, working on software, data centers, networks and services. A common theme I've seen is enterprises realizing they need to allocate more resources to innovation to be successful. This often starts with controlling costs, but cloud initiatives remain a top priority, and other major efforts, such as analytics, are tightly coupled to cloud strategies. Those strategies are rapidly evolving for a multi-cloud world.
Despite the rapidly growing interest in multi-cloud, most of the discussions miss its true driver: innovation. There are a number of other reasons that cause organizations to pursue multi-cloud, or believe they should: risk mitigation, triaging workload placement, redundancy, or survivability and avoidance of vendor lock-in, among others. None of these are unworthy reasons, but they distract from the story of innovation in the cloud.
A History Of Innovation
Historically, the story of the cloud was one of infrastructure. The major public cloud providers matched each other with instance types, network connectivity options and solutions for block and object storage. They expanded across the globe with new regions. They offered platform-level services, allowing users to easily spin up database instances, connect services with messaging and so on. Cloud-powered services such as relational database instances and load balancers emerged, which are also typically closely tied to infrastructure.
As the cloud matures, however, the story of innovation is shifting to services that are further up the stack. The large public cloud providers are directly tackling problems such as machine learning, big data, internet of things (IoT) applications and even developer tools integrations. This is the challenge for organizations adopting cloud. Many do understand that their cloud strategy is an investment in competitiveness. A simple question to ask is: if every major cloud is rapidly innovating, will we be OK with the innovations of one cloud provider, or do we need access to the innovations of each provider to be competitive?
These innovations have made powerful capabilities simpler, automated and egalitarian. The phrase “anyone with a credit card” has nearly become the rallying cry of cloud evolution. Developers, and even organizations as a whole, will be drawn to solutions that enable them to get their jobs done faster. Even now, people might turn to Amazon Redshift to process data and then pipe it into Microsoft's Power BI in Azure to visualize it. This type of use case is a mere indicator of what's to come in the multi-cloud wave.
So how do you plan for the multi-cloud?
- Have a plan for managing your data. Data management is one of the largest challenges facing IT leaders today. How can you collect, secure, store, process, visualize and (perhaps) eventually retire data? Comprehensively addressing concerns of locality, ease of access, security, privacy, durability and accessibility across your IT initiatives will require you to take a broad view and plan ahead for your digital initiatives. Moving data ad hoc over long distances isn't feasible as terabytes grow to petabytes.
- Plan for integration with cloud services from the start. It's a hybrid- and multi-cloud world, and it is highly likely that some data will remain on-premise, a great deal will land in colocation facilities and much will be ensconced in cloud services. Some have observed that workload placement will be a key driver of digital infrastructure delivery. It is critical to foresee how applications, their associated data and unstructured data lakes cooperate in a hybrid- and multi-cloud environment. There is no one-size-fits-all solution, but the prescription to plan now for how you move, store and access data when you need to integrate disparate environments is a constant.
- Ensure your strategy is multi-cloud ready. It's not enough to integrate with one cloud, if one cloud is not enough. The breakneck innovation of cloud providers will help ensure that, even if the other factors do not.
- Plan holistically for applications. An application is more than just data; applications are complex and require a variety of services. The fundamentals of "compute, network and storage" are the underpinnings, even in a microservices world, but when applications can stretch services across multiple clouds, you have to plan for the infrastructure resources and the application services, whether your application runs at the edge, in an enterprise data center or across multiple clouds. Future-proofing requires considering the flexibility of your data center today, in a multi-cloud world.
- Don't wait for developers to solve this for you. If shadow IT is over 30% of IT spend, but most data center closures are yet to happen, we can conclude that a great deal of disruption to where data lives has yet to occur -- and developers will do what they need to in order to ensure their applications work and evolve. If you force them into a do-it-yourself situation, then you will have fragmented results instead of a strategically sound outcome.
The time to begin planning for the multi-cloud world is now. Many IT leaders might love to turn back the clock on cloud to get a jump on the shadow IT adoption, implement a stronger strategy and better governance, and coordinate their organizations better. The good news is that you have the time to do this with multi-cloud. But make no mistake: The clock is ticking.
